|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 8/29/2017 Posts: 278
|
Question: Got this from SWMGamers: Is this still the rules for ranking players that are tied on points in a tournament? Also, what exactly do steps 3 through 5 mean? :
In the event of multiple players being tied for match points the following shall be used to determine the final standings: 1. Total Wins-The player with the most wins. 2. Head to Head Matchups – The player who has won any head to head matchups would be ahead of the loser in a head to head matchup. 3. SOS 1- The players strength of schedule minus the player's worst win. 4. SOS 2- As 3, factoring only the defeated players. 5. SOS 3-The players full strength of schedule 6. Roll Off-tied players will roll a D20, highest roll placing above the rest.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
SOS = Strength of Schedule. I'm pretty sure you'd calculate strength of schedule by looking at the combined record of the player's opponents. The difference between the three (steps 3, 4, and 5) is how many of each player's opponents are taken into consideration.
Step 3 ignores the worst win - in this step players are not penalized for getting the bye
Step 4 counts only the opponents a player defeated - in this step players are not rewarded for being lucky enough to play (but lose to) the #1 player
Step 5 counts everything
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 7/9/2008 Posts: 4,729 Location: Chicago
|
FlyingArrow wrote:Step 3 ignores the worst win - in this step players are not penalized for getting the bye
Actually I believe the bye only counts as a 3pt win but does not factor into SoS at all. This step is more about not punishing a player who was unlucky enough to play the opponent who does t win a single game.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
TimmerB123 wrote:FlyingArrow wrote:Step 3 ignores the worst win - in this step players are not penalized for getting the bye
Actually I believe the bye only counts as a 3pt win but does not factor into SoS at all. This step is more about not punishing a player who was unlucky enough to play the opponent who does t win a single game. I guess that leads to 2 important questions, then. 1. Does SOS count opponents' total wins, or does it count opponents' combined winning percentage? 2. Does the Bye count as an opponent with 0 wins and 0 losses, or does it count as an opponent that has lost every game? (moot point if the answer to question #1 is counting wins) I was thinking the answers were winning percentage and the Bye counts as an opponent with a bunch of losses, but I'm probably wrong on one or the other.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 4/30/2008 Posts: 2,093
|
so lets do an example. you play 5 opps and win 3 games and lose two. Opps record in parens.
rd 1 W (2-3) rd 2 L (4-1) rd 3 W (0-5) rd 4 L (3-2) rd 5 W (1-4)
So TB 3 SOS minus worst win (the guy who went 0-5) is 10-10 TB 4 as 3 but factoring only defeated players (so no 4-1 or 3-2 losses or the 0-5 win) 3-7 TB 5 Full SOS counts everyone 10-15
Also if I recall, the way HTH works is only if the player has played everyone else. So if A,B,C,D all end tied going to step 2, A is only ahead if it beat all of the others. And once you determine one spot I believe you restart the loop. So if D has lost to all A,B,C, then that player would be the lowest of those 4, and you would restart the TB process for A,B,C.
But just because A beat B HTH doesn't mean that A is ranked ahead of B unless the TB is just between those two. Because it is possible that D beat A but lost to B creating an infinite loop. So HTH doesn't come into play very often, because it requires a very specific set of circumstances.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 8/29/2017 Posts: 278
|
Thanks for the help guys! Richard is looking for the correct answer for his tournament parring/tracking app. If we could figure out a very accurate answer that would really help him out.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 7/9/2008 Posts: 4,729 Location: Chicago
|
How byes are calculated are really the crux of the issue.
If I go 3-2, and one is a bye, I believe my SoS is all opponent's records added and then divided by 4 (not 5), thereby the bye is discluded altogether.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
It makes sense that the bye isn't counted. I just hadn't ever been in a situation where it mattered so I didn't know.
According to Jason, you calculate SOS by adding the wins and losses for the relevant opponents and take the winning percentage. (And the bye is a 0-0 opponent, so it isn't included.)
Adding up the wins and losses is slightly different from what you said, Tim, in the case where someone drops out and hasn't played the same number of games as other players. If everyone played the same number of games, it wouldn't make a difference, though.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 4/30/2008 Posts: 2,093
|
Here is another example. Again going 3-2 (although this time with a bye and a player who doesn't play all 5 rds)
W (bye) W (1-2 drop) W (4-1) L (5-0) L (4-1)
So
3 SOS minus worst win (1-2 drop is worst win) so would be 13-2 4 SOS same as 3 but only defeated opponents. Only beat one other player so 4-1 5 SOS full schedule 14-4
Again these are Tie Breakers for the SOS section.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
For worst win, is the "worst win" the opponent with the lowest winning percentage or the opponent that hurts your SOS the most? Only relevant when someone drops out.
Opponents: R1 W 0-1 (0%) R2 W 1-4 (20%) R3 W 4-1 (80%) R4 L 5-0 (100%) R5 L 4-1 (80%)
3. SOS minus R1 opponent (0-1) is 14-6 (70%). SOS minus R2 opponent (1-4) is 13-3 (81.25%).
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 7/9/2008 Posts: 4,729 Location: Chicago
|
I always thought The record of the player you beat with the most losses is dropped (not percentage). This way a person dropping doesn't hurt you as bad as a person continuing on and continuing to lose
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
TimmerB123 wrote:I always thought The record of the player you beat with the most losses is dropped (not percentage). This way a person dropping doesn't hurt you as bad as a person continuing on and continuing to lose "Most losses", "worst winning percentage", and "the one that harms your SOS the most" are 3 different criteria - and all are valid interpretations of "worst win". They are only different if someone drops out of the tournament early, but zeroing in on the right one would be important for whoever is writing the software. Here "worst winning %" differs from "Most losses" and "most harmful" by a lot: R1 W 0-1 R2 W 1-4 R3 W 4-1 R4 L 3-2 R5 L 5-0 Drop Worst winning % (R1): 13-7 (65%) Drop Most losses / most harmful (R2): 12-4 (75%) Here "most losses" and "most harmful" differ: R1 W 0-2 R2 W 2-3 R3 W 3-2 R4 L 3-2 R5 L 3-2 Drop Most losses (R2): 9-8 (52.9%) Drop Most harmful / worst winning % (R1): 11-9 (55%)
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 8/29/2017 Posts: 278
|
Is this something we do not have defined previously? To this depth I mean...
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 4/30/2008 Posts: 2,093
|
Darth_Frenchy wrote:Is this something we do not have defined previously? To this depth I mean... I do not know what criteria our current pairing software uses or how the old excel program handled it either. And no, I don't believe it is defined anywhere what "worst win" means.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 8/29/2017 Posts: 278
|
Okay, do we know who decided on the criteria? Is this something the rules committee and/or the balance committee should figure out?
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 8/29/2017 Posts: 278
|
Okay, Urbanjedi said
"ok some digging. intent was to eliminate the win that hurt sos the most." Until further notice that is what I am going to go with. Thank you all for your help!!!!
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
Darth_Frenchy wrote:Is this something we do not have defined previously? To this depth I mean... As long as everyone sticks around for the whole tournament, all 3 interpretations are identical, so probably hasn't been a reason to dive that deeply.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 11/12/2012 Posts: 90
|
Wow guys! Thanks for the discussion on this topic! so SOS is calculated to be a percentage value, and then the "tied" players are distinguished by this percentage. I think as long as we show what the distinguishing factor was when ranking the tied players in the app, then it will avoid confusion/discussion/arguments ;) i.e. Bob 9pts > SOS = 59% , James 9 pts > SOS = 48%
for step 3 I can start with excluding the win against the opponent with the most losses, and test it. I wonder if these 3 options should be a "Tournament Configuration", and you choose which one you want.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
DarthRattlehead wrote:for step 3 I can start with excluding the win against the opponent with the most losses, and test it. I wonder if these 3 options should be a "Tournament Configuration", and you choose which one you want.
If you want the software to be flexible enough to work for multiple games/rules, sure. But for the SWMinis floor rules, I think you're safe just going with dropping the one that hurts the winning percentage the most.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 11/12/2012 Posts: 90
|
cool. ta.
|
|
Guest |