RegisterDonateLogin

Stacks

Welcome Guest Active Topics | Members

Our not-so-great tournament scoring system Options
TimmerB123
Posted: Saturday, April 25, 2020 11:02:52 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
This started on another thread - but it deserves it's own.

Here are the results after Swiss for the Vassal Regional:

spryguy1981 wrote:
Standings

*Spry 4-1 13pts
*thereisnotry 4-1 12pts
*FlyingArrow 4-1 10pts
*Mando 3-2 9pts
*Randy 3-2 9pts
*Gandalf 3-2 9pts
*DarkDracul 3-2 9pts
*General Grievous 3-2 9pts
TimmerB123 2-3 9pts
Kezza 3-2 7pts
Darth O 2-3 7pts
UrbanJedi 2-3 6pts
Caedus 2-3 6pts
Dr Daman 2-3 5pts
Kiki 1-4 4pts
UrbanShmi 1-4 4pts

I'm glad I didn't slip into the top 8, because I did not deserve to. I literally didn't win a single game, but ended up barely not making it. I got a bye and a concession/drop, and three 1pt losses. Had any one of 5 players not gotten a point somewhere along the way - I would have made it.

I think this is as clear of an illustration as to why 1 point losses are, and have always been, a bad idea for competitive play.

Good for encouraging children when they are first learning to play? Sure. You lost but here's a point for trying! But it really screws things up competitively.

Ultimately it should all be decided by record first and tournament points second, then it wouldn't be nearly as much of an issue.

FlyingArrow wrote:

I second this. For a long time I thought record came first and points as a tiebreaker. I was pretty shocked to learn that points were the primary ranking. Or if record isn't the primary critieria, at least make the spread bigger between points for a loss and for a win: 5pts for full win, 4pt win if you don't get 200, 1pt for 100pts in a loss, 0pts if you don't get to 100.
TimmerB123
Posted: Saturday, April 25, 2020 11:13:41 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
Yes, I had 3pts for a concession/drop, which isn't that common in the final round, but I also got 3pts from a bye, which is very common.

We must award max points ts. (in this case 3) for a bye or an opponent conceding/dropping. Can't punish a player for something 100% out of their control. That is no debate.

What is contested, is:

A) whether or not the first criteria should be record, or tournament score

B) should the tournament scoring be 3/2/1, 3/2/0, 5/4/1, etc.


I personally have no problem with tournament score being a tiebreaker. In fact it's a great one. A tiebreaker for players with the same record.

Currently the criteria for placement is:

A) Tournament score
B) Record
C) Head to Head
D-F) Mutliple strength of schedule measures

I propose:

A) Record
B) Head to head
C) Tournament score
D-F) Mutliple strength of schedule measures.

and then 3/2/1 is fine


And if Tournament Score has to stay #1 (and I don't see why it does, but . . .) Then I agree with TJ's plan of 5/4/1
thereisnotry
Posted: Saturday, April 25, 2020 9:29:32 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 4/29/2008
Posts: 1,679
Location: Canada
The whole reason that we adopted the 3-2-1 scoring system (developed by the NZ group) in the first place was to put a stop to the regular pattern of 56-48 final scores after an hour of play. People weren't playing to reach the build total (ie, 200pts), but were instead camping in gambit and sniping single pieces here and there, in order to force the opponent to have to charge into the deathtrap on your side of the board.

There were issues with slow play, analysis-paralysis, stalling (or at least combat avoidance). Commanders sitting in the back, lockouts, etc. These patterns were taking place across the whole community, precisely because they were encouraged by the scoring system we had at the time: the 2-0 scoring system. That system treated a 21-20 tiebreaker win as equal to a 200-75 full win. And so those terms--"tiebreaker" vs "full win"--started to be used, and people started to catch on and agree that a "win" due to tiebreakers did not deserve the same weight in a tournament setting as a full win where players had fully engaged in combat.

As a result of the adoption of the 3-2-1 scoring system, we started to hear judges saying "200 or 2, no exceptions" when outlining the rules for the tournament. That is, they weren't going to allocate a full win (ie, a 3pt win) unless you reached the full 200pts. Therefore, the onus was now on the players, to play fast enough to finish the game within an hour. No more taking 3 minutes to decide to spin a Mouse Droid! Miraculously, the tactics and players that relied on non-engagement and hiding began to find less success in tournament play, and the community learned to play more quickly.

All of this is simply to say that we as a community had a very good reason for adopting the 3-2-1 scoring system: slow play was almost everyone's biggest NPE. But why the "1" in the 3-2-1 system? Because if someone sees that he's got a bad matchup, what is to prevent him from turtling all of his guys in the back corner of the map, thus preventing his opponent from gaining a full win via combat? The attitude of, "Well, if I don't stand a chance of winning this match, then neither will you!" was something that we had to avoid. Furthermore, it incentivized those who faced a bad match-up to play their best anyway, because they may still be able to salvage a point out of their loss.

The end result was that someone who was 3-1 after 4 rounds of play could end up with 9 or even 10 pts, and make the top cut, even though someone else was 4-0 with 8pts (no full wins). The question is, which player played better and therefore deserved to advance? The very clear consensus across the community was that the player who played with full engagement was the one who should be in the playoff rounds.

Ok, that review of history is enough for now.


To answer the 5-4-1 system, I think it comes down to math: A full win in the 3-2-X system is a 50% better than a tiebreaker win (ie, it's one-and-a-half times more valuable than a tiebreaker)...by contrast, a full win in the 5-4-X system is only 25% better than a tiebreaker win. If we went with the 5-4-X system, it would be totally possible for someone with a series of tiebreaker wins (ie, final scores of 46-42, 20-21, etc) to make the playoffs. And suddenly, we'd be back in the place we were before, where non-engagement and slow play dominate the tournament scene.

So yes, I do think that Tournament Score needs to be the primary criteria for tournament ranking, followed by W-L record. Whether we stick with 3-2-1 or move to 3-2-0 is something that I could go either way on.

That's how I see it.
CorellianComedian
Posted: Saturday, April 25, 2020 9:57:48 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 8/30/2014
Posts: 1,048
Any reason 5-3-1 is out of the picture?
General_Grievous
Posted: Sunday, April 26, 2020 8:56:02 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 1/8/2010
Posts: 3,623
Thereisnotry convinced me, you have my bow.
TimmerB123
Posted: Monday, April 27, 2020 2:33:35 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
thereisnotry wrote:
The whole reason that we adopted the 3-2-1 scoring system...


I think time has faded your memory on some details. While a lot of what you said is true, but not related to 1pt losses at all.

thereisnotry wrote:
3-2-1 scoring system (developed by the NZ group)

True

but
thereisnotry wrote:
in the first place

false.

We developed the 3-2-0 system to combat what you speak of. The 3-2-1 NZ system came years later.


thereisnotry wrote:
As a result of the adoption of the 3-2-1 scoring system, we started to hear judges saying "200 or 2, no exceptions"

again false. 200 or 2 came about with 3-2-0 system, years before 3-2-1 was implemented.

thereisnotry wrote:
Miraculously, the tactics and players that relied on non-engagement and hiding began to find less success in tournament play, and the community learned to play more quickly. etc etc etc

True - but this happened after 3-2-0 was implemented, NOT after 3-2-1 which came years later.

You continue to give reasons why 3-2-0 happened, and I think there was only 1 sentence correlating directly to adding 1pt losses. (unfounded fears of players running to hide in a corner.) This has always been possible, and has always been illegal. Call a judge. It has happened very rarely in my experience. A equal argument could be made on the other side about players running headlong into their opponent to gift them with a 3pt win they didn't "earn". Seen that happen many times.

The grand irony is all of the arguments Trevor presented above favor the 3-2 system, but actually are CONTRARY to 1pt losses. I'll get into this in a bit.

But first, let me clear some things up, because I was literally there, live in person, in the room, when a lot of these decisions were made.

some accurate history:

2007 GenCon championships. A certain player went 5-2 in the 130-some player Gen Con championships (biggest SWM tournament ever). That player did not make the top 8, but 2 players he beat did. 2 of his early opponents dropped, and that killed his strength of schedule, which was THE ONLY TIE-BREAKER at the time. Virtually everyone thought this was a bad criteria as a first tiebreaker, so they set out to come up with a better system. They also wanted to address the things Trevor talked about.

So 3-2-0 was implemented, and great emphasis was put on FINISHING GAMES. Not only did the tournament point system become the primary criteria, but the powers-that-be wanted to double down on emphasizing 3pt wins. In the tiebreaker system (at that time),
1. tournament points
2. record
3. number of 3pt victories
4. then head to head
5. then strength of schedule (SoS).

Now the funny thing is, number of 3pt wins (3rd tiebreaker) was moot when you already have tournament score and record above it. Simple math shows it. 4-1 with two 3pt wins = 10pts. It's impossible to have 10pts, 4-1 record and NOT have two 3pt wins (at the time). Even when this was pointed ted out, it was ignored, since it wasn't hurting anything.

As Trevor said (accurate description, but inaccurate timeframe and cause) " Miraculously, the tactics and players that relied on non-engagement and hiding began to find less success in tournament play, and the community learned to play more quickly".

It was really problem solved at that point.

There was some changes made to SoS after that, since we were still having ties, and there were other factors to consider such as getting beat by the undefeated player should not count against you as much as losing to the 1-4 player. However beating the 0-5 player should not could against you (not your fault you got paired) as much, and beating the 5-1 player should count well for you. We split SoS into 3 parts:
1. opponents win percentage, excluding the defeated opponent with the lowest percentage (not punishing you for getting paired up with the newb in round 1)
2. defeated opponents win percentage, excluding the defeated opponent with the lowest percentage. (Weighing only who you beat, while still also not punishing you for getting paired with the newb)
3. Total SoS

but keeping all the rest the same.


So THEN (again, YEARS later) 3-2-1 was implemented. It was fairly ingrained into the New Zealand group by then, and they were resistant to abandoning it. Some of us Yankees were for it, some where against it. In the end it was implemented to make sure we matched up with the then roaring SWM scene in NZ.

Incidentally - at the exact same time we also switched from 5pt for gambit to 10 points.


This was 2 changes at once, and any scientist knows you need to isolate variables to get the best data, so a lot is muddled after that.

A lot of semantics were played along the way (some of which Trevor spoke of). If you win a game 199-0, that is a tie (what?) but you get the "tie-breaker" (silly). 3pt win and 2pt win is fine thank you.

Then there were some ironic twists. More games "finished" (imo more to do with 10pt gambit than 1pt losses), but what does "finish" mean? The game as created was always meant to be "finished" when one team was completely destroyed. Period. Much fewer games "finished" in the true sense - they just reached 200 points, often due to abusing gambit (instead of lock outs - we now had lock-ins, lol). Now some people think this is better, some just think it reverses the problem. As a competitive player - if it's legal, use it. I'm not shaming anyone playing smart within the rules, including lock-outs or locking yourself in gambit to win. But as part of the design and leadership team of the community, I personally would rather see more games finished in the real sense, not due to gambit.

Another funny twist happened next. A player at GenCon missed making the finals after tying other players in tournament score, tying them in record, but having more 3pt wins. How is this possible?


Remember - it was tourney score, record, then most 3pt wins.

Well, apparently there was a lot of confusion after 3-2-1 came into play, and some things fell through the cracks.

Once 3-2-1 was implemented, that 3rd tie breaker finally MATTERED! You could be 4-1 with 10 points and two 3pt wins (and a 0pt loss), or you could be 4-1, with 10pts and only one 3pt win, but also a 1pt loss.

For years emphasis was put on finishing games, finishing games, FINISHING GAMES. Nothing more important than 3pt wins, it's both a major part of the 1st criteria for ranking, and the next most important thing after record!!

Except, after being told the 3rd criteria was moot for years (which it literally was with 3-2-0), the powers that be decided to drop it, IRONICALLY, when 3-2-1 was implemented - WHEN IT FINALLY WOULD MATTER.

Except the powers that be didn't tell everyone it was dropped. In fact it wasn't even updated on the forums (it still said 3pt victories was the 3rd criteria).

This was debated, and (mainly due to the player being disliked by a certain contingent of the powers that be) it was decided that a 2pt win (or tie-breaker as they tried to push) and a 1pt loss SHOULD be exactly the same as a 3pt win and a 0pt loss. COMPLETELY REVERSING THE STANCE THEY HAD BEEN ADAMANT ABOUT FOR YEARS!





What is undoubtably clear - the system prized playing fast above playing well. Of course both is the best, but the fact that you can win more games than someone but place lower is definitive proof of that.

Some people think that is good. Others do not.




I am fine with tournament record being first in 3-2-0, but I am not OK with it in 3-2-1, and the recent tournament results illustrate why.

As a game and as a community, we have evolved quite a bit. Stalling has not been in issue in years (since before 3-2-1), but EVEN MORESO now. We have a smaller community. We are all essentially friends. Overall we collectively play faster. We are not as cut-throat. We have intentionally designed sets to push more action, and that has been successful.



What I recommend, is thus:

1. Record. (I'm sorry, but playing fast should not trump winning.)
2. Head to head (this should be more important. I beat you, but we have a tied record. I should finish above you>
3. tournament score. (Record is VERY OFTEN tied. It is strong motivation to always finish games. H2H rarely comes into play)
4. # of 3 point wins. (again, prioritize FINISHING games, not rewarding losing players. {4-1, 10pts, two 3pt wins} IS BETTER THAN {4-1, 10pts, one 3pt wins}. Currently they are equal.
5. Strength of Schedule A, B, C (the same 3 breakdowns)



I think it's absolute hyperbole to say we would be right back where we started if we changed anything. In fact, I think it would be IMPOSSIBLE to go back there.

If you truly prioritize full wins, 1pt losses hurt the system.

I agree with Trevor when he talks about it coming down to math, I just disagree with his subjective opinions on how to judge the math.

Say a player gets a 2pt win. With 1 pt losses, they can literally just bum-rush their final opponent, knowing they will lose, simply to get to 100 (fairly easy to do if that is your only goal). Then that "made up" for your 2pt win (errr tie-breaker? That semantic propaganda is so silly).


So, in this case a player could go 4-2 with 10pts with two full wins, and lose to another player who NEVER FINISHED A SINGLE GAME (4-2, 10pts, wins in SoS, four 2pt wins and two 1pt losses)

Trevor - over and over your arguments only apply to the 3-2 system, and say nothing about the 1pt.

(and side point, yes - a full win should be 25% better than a (currently) 2pt win. 50% is insane. We can talk about that later)

You use extreme examples. Let me use the other extreme.

How about this for each player's in-game scores:
W 199-0, W 199-0, W 287-0, W 264-0, L 99-101, L 99-100 (4-2, 10pts)
vs
W 10-5, W 22-3, W 56-55, W 5-0, W 100-286, 101-255 (4-2, 10pts)

You're telling me there isn't an absolutely obvious player that should be ranked higher? In current system, this could be decided by strength of schedule, and the 2nd player could win that. LUDICROUS!!!




So - in conclusion, if you are truly for games finishing, you should be AGAINST 1pt losses, because it allows people who don't finish games to catch up.



So - If tournament score needs to be the primary criteria, (and I believe it does not) then 3-2-1 needs to change.

easiest would be to drop 1pt losses, but if that has to stay, then it should be 7-4-1.


Those numbers seem random, right?

Let me explain why:

A 1pt loss + a partial win CANNOT = a full win. It is explained in depth why above.

If we truly prioritize FINISHING GAMES, then we have to give greater reward to FINISHING GAMES.

If you don't finish your wins, virtually impossible to catch up with losses.

However, it still essentially acts as a tiebreaker for those with the same record and # of 3pt wins.


so this is why 5-4-1 won't work. (4+1 = 5+0, AND IT NEEDS TO NOT!)

5-3-1 won't either. It would let a player who doesn't finish games "catch up" with just two 100pt losses. It should be harder than that.

6-4 is exactly the same as 3-2 and we are trying to do better


So this brings us to 7-4-1


Let's look at some math. (Same example as above)

W 199-0, W 199-0, W 287-0, W 264-0, L 99-101, L 99-100 (4-2, 28pts)
vs
W 10-5, W 22-3, W 56-55, W 5-0, W 100-286, 101-255 (4-2, 18pts)

Now that is more like it!!!


The player who did not finish any games ended up much lower, rightfully so.

Arguing for 3-2-1 is arguing for the player who didn't finish a game the chance to go to the finals.

Added benefit is there will be far fewer ties, and FULL WINS will be the most heavily weighed by far.



So - if I am in the minority of people that think record should be paramount (and that is by no means clear these days, but I am allowing for that possibility), then I can accept that.


However, I think that the vast majority of players want to prioritize FINISHING GAMES. Therefore should also be AGAINST 3-2-1. (I am sure some people haven't thought it all the way through).

barring dropping 1pt losses altogether (again, I may be in the minority thinking we should), then we MUST raise 3-2 (and I like 7-4)



Then I would recommend this:

1. tournament score (7-4-1 system)
2. Record.
3. Head to head
4. # of 3 point wins. (This currently doesn't exist, and it needs to come back if we have 1pt losses)
5. Strength of Schedule A, B, C (the same 3 breakdowns)
TimmerB123
Posted: Monday, April 27, 2020 2:59:03 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
General_Grievous wrote:
Thereisnotry convinced me, you have my bow.


Might want to reconsider with corrected info above
TimmerB123
Posted: Monday, April 27, 2020 4:09:32 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
So we can make up crazy scenarios all day - but here's how 7-4-1 would work in a REAL WORLD test

After Swiss in the vassal tournament:

*Spry 4-1 13pts.
*thereisnotry 4-1 12pts.
*FlyingArrow 4-1 10pts.
*Mando 3-2 9pts.
*Randy 3-2 9pts.
*Gandalf 3-2 9pts.
*DarkDracul 3-2 9pts
*General Grievous 3-2 9pts
TimmerB123 2-3 9pts.
Kezza 3-2 7pts.
Darth O 2-3 7pts
UrbanJedi 2-3 6pts.
Caedus 2-3 6pts.
Dr Daman 2-3 5pts.
Kiki 1-4 4pts.
UrbanShmi 1-4

would end up like THIS if we were doing 7-4-1

*Spry 4-1 (7, 7, 1, 7, 7) = 29pts
*thereisnotry 4-1 (0, 7, 7, 7, 7) = 28pts.
*FlyingArrow 4-1 (7, 7, 4, 4, 0) = 22pts.
*Mando 3-2 (7, 0, 7, 7, 0) = 21pts.
*DarkDracul 3-2 (0, 0, 7, 7, 7) = 21pts.
*Randy 3-2 (0, 7, 7, 1, 4) = 19pts.
*Gandalf 3-2 (7, 4, 0, 7, 1) = 19pts.
*General Grievous 3-2 (1, 4, 7, 0, 7) = 19pts.
TimmerB123 2-3 (1, 1, 1, 7, 7) = 17pts.
Kezza 3-2 (7, 4, 4, 0, 0) = 15pts.
UrbanJedi 2-3 (7, 0, 0, 7, 0) = 14pts.
Caedus 2-3 (7, 0, 0, 0, 7) = 14pts.
Darth O 2-3 (4, 0, 1, 1, 7) = 13pts.
Dr Daman 2-3 (0, 7, 0, 4, 0) = 11pts.
Kiki 1-4 (7, 0, 1, 0, 0) = 8pts.
UrbanShmi 1-4 (0, 0, 7, 1, 0) = 8pts.

Not too many position changes, but DarkDracul jumped from 7th to 5th (because he GOT MORE FULL WINS!), and Randy and Gandolf both got bumped down 1. (If this were GenCon and cutting to a top 6, DarkDracul would have been screwed with the current system!)

Lower in the bracket Darth O drops a couple spots because he only got 1 full pts win, but stays above Daman because he got 1pt losses. (this all seems to me as it should be)

One thing that didn't change - I still placed above Kezza, due to fairly extreme results. Kezza only got full points once, and 0 in his losses. I got full points in both wins, and 1 point in all 3 losses. Had he just finished 1 more of his wins, he'd be above me (rightfully so - he won more).



So - does this change a lot? No, but it is more accurate if we, as a community, really want more games to finish. Prioritize FULL WINS.

I think this would encourage people EVEN MORE to play to completion
Darth O
Posted: Monday, April 27, 2020 5:10:03 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 6/30/2009
Posts: 1,389
Location: New Zealand ( kind of by Australia)
I really like the 7-4-1 proposition now. Well argued, Tim! You have my vote if there is one.
TimmerB123
Posted: Monday, April 27, 2020 5:29:14 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
Darth O wrote:
I really like the 7-4-1 proposition now. Well argued, Tim! You have my vote if there is one.


Huzzah! (I don't think it will work like that, but glad to have the support!)
FlyingArrow
Posted: Monday, April 27, 2020 8:47:12 PM
Rank: Moderator
Groups: Member , Moderator

Joined: 5/26/2009
Posts: 8,407
I like it but 3pt wins are so ingrained into our lexicon that I think we need to go with:

3pts for full wins,
1.71pts for a partial win,
0.14pts for a 100pt loss, and
0pts for a loss below 100pts.

Works the same as what Tim proposed but is simpler because we can still talk about 3pt wins.
kezzamachine
Posted: Tuesday, April 28, 2020 12:01:47 AM
Rank: Moderator
Groups: Member , Moderator

Joined: 9/23/2008
Posts: 1,467
Location: Lower the Hutt, New Zealand
Speaking of 3pt wins, I've uploaded heaps of videos of my music onto YouTube. Really think watching them all and subscribing to me ("kezzamachine") would give you all a more well rounded view of the 3-2-1 system.
kezzamachine
Posted: Tuesday, April 28, 2020 12:15:01 AM
Rank: Moderator
Groups: Member , Moderator

Joined: 9/23/2008
Posts: 1,467
Location: Lower the Hutt, New Zealand
Okay - so in our first 2-3 tournaments (back in 2008), we used to score based on in-game points killed(!) as we wanted to reward action over the 2-0 game-score that I think was prevalent around the time (from memory). After a couple of these, we heard about the new-ish 3-2 system and changed over to that. Some time over the next few years (but can't remember when), Bevan (from Hawera, NZ) and I developed the 3-2-1 as we wanted to (as you know) encourage the intent to fight, no matter what the state of the game you were playing - you could still get something out of it!

Also... something of note which surprised me when I logged on to the Bloo today: I was actually moderately shocked that I had slipped from playing on the Top Table in Round 4 to being 10th!!! I can absolutely see how it happened - no complaints! - but I found it fascinating that I had dropped so incredibly far in two games! Wowser. The 5th round hurt me!

Peace out. Subscribe.
TimmerB123
Posted: Tuesday, April 28, 2020 3:15:38 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
kezzamachine wrote:
Okay - so in our first 2-3 tournaments (back in 2008), we used to score based on in-game points killed(!) as we wanted to reward action over the 2-0 game-score that I think was prevalent around the time (from memory). After a couple of these, we heard about the new-ish 3-2 system and changed over to that. Some time over the next few years (but can't remember when), Bevan (from Hawera, NZ) and I developed the 3-2-1 as we wanted to (as you know) encourage the intent to fight, no matter what the state of the game you were playing - you could still get something out of it!

Also... something of note which surprised me when I logged on to the Bloo today: I was actually moderately shocked that I had slipped from playing on the Top Table in Round 4 to being 10th!!! I can absolutely see how it happened - no complaints! - but I found it fascinating that I had dropped so incredibly far in two games! Wowser. The 5th round hurt me!

Peace out. Subscribe.


I should make clear that my opposition to 1 point losses is no personal indictment on you or the New Zealand community. You guys rock. Sincerely.

Your passion for the game and the community that was created around it are genuinely inspiring.

For me it’s the principal. I completely understand the motivation behind it. And, as I hope is clear by my far-too-long post, even though I personally don’t think it’s necessary, I am willing to keep it - so long as the “3-2” part is adjusted to be worth appropriately more.

When I really dove into the math, 50% of a partial win and 33% of a full win is far too much. As I pointed out, giving that high of a percentage allows players with zero full wins to do BETTER in tournaments.


TimmerB123
Posted: Tuesday, April 28, 2020 3:20:31 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
FlyingArrow wrote:
I like it but 3pt wins are so ingrained into our lexicon that I think we need to go with:

3pts for full wins,
1.71pts for a partial win,
0.14pts for a 100pt loss, and
0pts for a loss below 100pts.

Works the same as what Tim proposed but is simpler because we can still talk about 3pt wins.


You really think that's easier!?! LOL


Semantics wise - I want to promote these terms for our permanent usage:

Full win

Partial win


Simple, clear, and effective regardless of points associated. It also has the subtle verbal encouragement to finish games (without the ridiculous semantic propaganda of one being winning by “tie-breaker”. There’s definitely a reason that never really caught on)
kezzamachine
Posted: Tuesday, April 28, 2020 8:16:25 AM
Rank: Moderator
Groups: Member , Moderator

Joined: 9/23/2008
Posts: 1,467
Location: Lower the Hutt, New Zealand
TimmerB123 wrote:

I should make clear that my opposition to 1 point losses is no personal indictment on you or the New Zealand community. You guys rock. Sincerely.


To be clear in return, I was taking a trip down memory lane. I only really want subscriptions!
shmi15
Posted: Wednesday, April 29, 2020 1:23:26 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 4/19/2010
Posts: 1,290
Why not score it based on how many points you score. Like... if you win 3 games. And you accumulate 75, 213, and 100 points, why can't your tournament score be 375? That would make the final round extremely interesting, and you'd almost never have to go to tie breaker 3, which would be head to head.

Make 200 the max points possible.

System would be

Tourney Record
Total points accumulated
Head to head
TimmerB123
Posted: Wednesday, April 29, 2020 10:25:20 AM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
shmi15 wrote:
Why not score it based on how many points you score. Like... if you win 3 games. And you accumulate 75, 213, and 100 points, why can't your tournament score be 375? That would make the final round extremely interesting, and you'd almost never have to go to tie breaker 3, which would be head to head.

Make 200 the max points possible.

System would be

Tourney Record
Total points accumulated
Head to head


I absolutely understand the impetus behind this idea. However, in my opinion there are just too many "moving parts". It's far too easy to miss something, record a score inaccurately, etc. Also that would allow for greater swings with shenanigans. I know I'm beat so I just go Kamakazee to score a bunch of points, even though I have no chance of winning with such a method. I'm not saying this would happen, but it easily could.

Additionally, it advantages certain squad types.

Scenario 1: A big beefy 100pt character is all that's left and down to 10hp. All action took place outside gambit, and both players went straight to battle. Sounds like a gave of a lot of engagement to me, however, the losing player would only get 100pts.

Scenario 2: A game where the winner has 150pts of their figs left on the board, but the loser got 6 rounds of gambit and killed 50pts worth of characters. They out activated and just shot a character into gambit at the end of the round 6 times, and managed to kill a few other characters. Losing player receives 110pts

In this scenario - the player in the first scenario would score less than the player in the second scenario, even though it's clear there was much more engagement and attempting to win in the first scenario.


I wouldn't mind it that much if it were a tie-breaker further down the list, but I don't see it feasible as the first or second criteria.
thereisnotry
Posted: Wednesday, April 29, 2020 12:28:47 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 4/29/2008
Posts: 1,679
Location: Canada
Ok, now I understand more clearly where you're coming from, Tim. You're not against the 3-2 scoring system, but you're against the 3-2-1 scoring system. I didn't understand that properly when I first read your posts (it was 2:00 am, after the conclusion of the Chicago Regional, so I was more than a little tired).

Looking back and reading again, I can now see where I misunderstood you. More specifically, I now see the main phrase that I tripped over:

TimmerB123 wrote:
Ultimately it should all be decided by record first and tournament points second, then it wouldn't be nearly as much of an issue.


In other words, I misunderstood this to mean that you wanted to go back to the W-L system. That's why I went into the reasons why we established the 3-2-X system in the first place. If you re-read my post, you'll see that just about the whole thing is talking about the importance of encouraging full wins, rather than tie-breaker wins.

And yes, the phrase "tie-breaker win" was one of the descriptions that I remember being used back in the day while these discussions were brewing, because the people using that phrase were trying to get across to everyone that a 56-49 win is not equal to a 200-X win. So yes, it was absolutely a phrase in use at the time. Now we simply talk about 3pt and 2pt wins, which is totally fine by me.

As for the history, I'm not surprised that I didn't remember it as clearly as you did. I was just someone playing the game at that time, and as I recall I wasn't a leader or a decision-maker, and so I probably wasn't consulted for the more back-room or high-level discussion on the topic. I just remember the discussions on the boards about slow play and the need for a 3-2 system. So I apologize for the inaccuracies in my attempted re-telling of the history of this discussion.


If I had properly understood your posts from the start, I probably wouldn't have even taken the time to write a reply, since I could go either way on it. And actually, I said that in my post:
thereisnotry wrote:
Whether we stick with 3-2-1 or move to 3-2-0 is something that I could go either way on.


After reading your long post about why the 3-2-1 system isn't ideal, I think I agree with you: 3-2-0 would be better. (Is that, indeed, what you're arguing for?) I think you make a good point, that it should be very difficult (or even impossible) for a player with less full wins to place higher than one who has more full wins.
If that's the case, then you have my "vote" and my axe.
TimmerB123
Posted: Wednesday, April 29, 2020 6:51:48 PM
Rank: Advanced Bloo Milk Member
Groups: Member

Joined: 7/9/2008
Posts: 4,729
Location: Chicago
Trevor, without a doubt my post was lengthy and wandering. Easy to see why it could be misunderstood. On a personal level, I do think record should be paramount with full wins coming in as a tie-breaker. BUT I concede that it would be too big of an ask (so to speak) and would have strong resistance.

So, noting that - I'm focusing more at looking at 1pt losses:


EITHER we need to increase value of partial wins and full wins, thus essentially making 1pt losses factor in, but to a much smaller degree (which I am fine with)

-OR-

Drop the 1pt wins altogether.


It seems that a lot of people like 1pt losses, so I think the simplest fix is my 7-4-1 suggestion


7pts = full win (200pts+)

4pts = partial win (>200pts)

1pt = high score loss (100+pts)

0pts = loss (>100pts)


If we truly want to emphasize higher scores - we could add more

Heck - I might even go 1 step further and say:

5pts = high score partial win 100-199pts

and

4pts = low score partial win (>100pts)

so it's really a 7-4 system, but you get an extra point for partials and losses over 100pts.


I guess this is my favorite proposal (factoring in where I am possibly not in the majority and/or there might be high resistance)

7pts = full win (200+ pts)

5pts = high score partial win 100-199pts

4pts = low score partial win (>100pts)

1pt = high score loss (100+ pts)

0pts = loss (>100pts)



As for the vernacular - I recommend we use the above terms instead of numbers. That way no matter what the system is - it's clear. Full Win, Partial Win (high score/low score if we adopt that), High Score Loss.

Also - we wouldn't have to say 7-5-4-1. We would just call it The 7 point system



Other recommended changes. Seeing as we are currently prioritizing tournament score above record - it has never really made sense that the strength of schedule uses record. It should use tournament score as well.



so - in conclusion:

The TOURNAMENT RANKING CRITERIA would be thus:

1. Tournament Points (using the 7 point system)

2. Record

3. Head to Head

(Note - if more than 2 players are tied at this level, one player must have beaten ALL other players to
be ranked higher here, or else it moves onto the next tiebreaker criteria. This is how it currently works.)

4. Number of FULL (7pt) Victories

5. Strength of Schedule

A) Average tournament score of opponents, excluding the defeated opponent with the lowest score
B) Average tournament score of opponents you defeated, excluding the defeated opponent with the lowest score.
C) Average tournament score of opponents who defeated you
D) Average tournament score of all opponents

6. Roll off


Users browsing this topic
Guest


Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

Main Forum RSS : RSS

Bloo Milk Theme Created by shinja
Powered by Yet Another Forum.net.
Copyright © 2003-2006 Yet Another Forum.net. All rights reserved.