Prize Calculation, or why did I wake up Murphy :)

Running Chess Tournaments
1

Rob Getty asked a question about prize calculation on Saturday and sure enough, Murphy was hanging around and decided to conjure up almost that same scenario. I hate it when Murphy is lurking about as it often makes Tim Just’s signature about the fine line between hobby and mental health even more profound… If I were Rob, I wouldn’t ask such questions when I am getting ready to run an event as it most assuredly woke Murphy up :slight_smile:

Below are the final results and prizes from a tournament Rob ran for the
Virginia Chess Federation this weekend. When Murphy reared his ugly head and was cackling in the background, Rob called seeking advice about calculating the below prizes.

2387 - 4.5
2181 - 4.5
2083 - 4.0
1987 - 4.0
1778 - 4.0
1946 - 3.5
1848 - 3.5
1833 - 3.5
1746 - 3.5
1628 - 3.5
1529 - 3.5
2034 - 3.0

1st - $450
2nd - $250
3rd - $150
Top X - $125
Top A - $125
Top B - $125
Top C - $125

1st-2nd easy division of $700 for $350 each
The people with 4.0 will split $150. $125, $125 for $133.33 each.

The question becomes which players bring in which prizes, as it then effects the remaining players waiting for money. I think it is correct to deal with the 4-1’s as a group with no consideration with what happens to the 3.5’s.

There are three possible solutions, with more if you get into the strange fractional prizes mentioned in the previous thread.

A.) 2083 brings in the 3rd prize, A/B Bring their class prize in, and the expert with 3.0 gets the unclaimed X prize
B.) 1987 brings in the 3rd prize, X/B bring in their class prizes and the three A players with 3.5 share the unclaimed $125 A prize
C.) 1778 brings in the 3rd prize, X/A bring in their class prizes and the two B players share the unclaimed $125 B prize
D.) Retire and go on a chess cruise :wink:

The above finish and prizes are real. He called me asking about how to award the three players with 4-1 and which prizes to give. There were players standing around arguing why this person or that person should be given the third place, most likley because they would benefit from the left over prize.

My answer was to start at the top and work my way down awarding prizes
downward. Since third place was “higher” than any other class prize, the expert brought that into the pool. The A/B players brought in their class prizes which left the Expert prize along with the other class prizes. For me, it was over and done with. It didn’t matter than the remaining expert had 3.0 and that there were Class A and B players above the expert with 3.5 points. So, my personal choice was solution A above.

I input everything into Swiss Sys (Rob used WinTD unfortunately, poor players must have suffered with nasty looking wall-charts all weekend) to see what its prize module would do. Interestingly, the pairings Swiss Sys came up with was exactly the same round for round except for a couple of the bottom boards in rds 3-5. This is a discussion for another thread on another day. (The windows features of Swiss Sys are just so much better than WinTD, which looks the same as it did a decade ago – oops, couldn’t resist :slight_smile:

Swiss Sys chose Solution C and an e-mail is off to Thad to find out why. Why is giving the 1778 the 3rd prize in the pool better than any other solution. Why aren’t the three A players equally deserving of splitting the A money if the A player brings in 3rd place? Why isn’t the philosophy of awarding prizes downward, with place before class and the X getting 3rd better? I really hoped Swiss Sys would back me up!

This would make a great test question, with emphasis on the justification.

Any comments? Is there a “correct” answer or is this just a matter of preference?

Michael Atkins

2

I agree with Swiss-Sys. The three players with 4 should divide the three “highest” prizes to which any of them are entitled. Place prizes are “higher” than class prizes, and X is “higher” than A/B, etc., so the three of them should divide 3rd/X/A. The problem you’re running into here is that trying to identify the player with the prize produces the seemingly paradoxical result of the B-player “winning” 3rd, but I think that’s just a matter of stating the problem incorrectly.

3

John,
What swiss sys did, when it produced the prize winners is to award third place to the B player, and then bring in X and A prizes to complete that three player pool. Since the prizes are the same for the class prizes, it didn’t matter which two class prizes are in there for money purposes. I think I see what you are saying here, that I brought the player into the pool rather than the prize, but that is part of the semantic problem in this situation.

In order for the prizes to be awarded 3rd, X, and A, Swiss sys had to assign third place to the B player, so that X and A were paid and the B prize left for B players. Definitely a paradox and interesting.

MA

4

Since the USCF rules order the class prizes, it is simplest to think of them as essentially:

Top X - $125.004
Top A - $125.003
Top B - $125.002
Top C - $125.001

(where roundoff gets your actual numbers). Looking at them in this way, it’s easy to see why Swiss-Sys produced the prize distribution that it did.

5

Lesson for the future: Make sure no two prizes are exactly the same.

Tim

6

Yes, I would also suggest using Under prizes, say something like:

U2200 $135
U2000 $130
U1800 $125
U1600 $120

Consider your original example, with the change of the 2083 to 1983. Then you would have the following distribution:

2034 - 3.0 $125 Top X

While lower rated players with better scores go home without a prize:
1946 - 3.5
1848 - 3.5
1833 - 3.5
1746 - 3.5
1628 - 3.5

7

I don’t have a problem with that, its seen all the time when experts and masters wind up with nothing after the place prizes are given out and players with a lot worse scores go home with money. If its based on class, then the competition is among class. The A and B players can blame their fellow class player who scored 4 for taking their money.

There are arguments for having under prizes but sometimes those are not fair to higher rated players. An improving 1700 player is eligible for U1800, U2000 and u2200 prizes and can easily sneak the money away from an expert who is only eligible for U2200 class prize. Having class prizes and not Under prizes helps to ensure money across classes ;“I can still will money in my class” instead of “That blasted B player had the tournament of his life and took the Expert money.”

You can make arguments for both systems, although I agree that having different prizes sometimes makes sense. It is hard to explain why a field of 5 experts warrants a $200 class prize when there is 20 B players fighting for $160. When I raise prize money with good turnouts, I’ll increase it in classes based on turnout, to reward the classes with better attendance.

MA

8

That strange fractional prizes become:
each of the 4.0-1.0 players get 1/3 of third place plus 2/3 of 1/3 of the sum of the three class prizes (since third is already included, only two more prizes can be brought in for the three players). 1/3 of third is $50. 1/3 of the sum of the three class prizes is $125, and 2/3 of that is $83.33.

That still results in the three 4-1 players getting $133.33, while leaving 1/3 of each class prize ($41.67) for the next group of players in each class (3 A players at 3.5-1.5 get $13.89 each, the 2 B players at 3.5-1.5 get $20.83 each, the lone C player at 3.5-1.5 gets $125 and the lone expert at 3-2 gets $41.67).

Martinak had a good explanation for why SwissSys did the distribution its way. You could always couple that with Tim’s comment and spring for an extra six cents in the prize fund to make it:
1st - $450
2nd - $250
3rd - $150
Top X - $125.03
Top A - $125.02
Top B - $125.01
Top C - $125

If the prizes are based-on then the change could be posted on-site after registration (during round two?). For that matter, if the prizes are guaranteed then I’d think the “massively ( :smiley: )” increased prize fund could still be posted. No grand prix point change though.

9

As one of the A players with 3.5, I am really glad I didn’t stick around for prize distribution! So what was the final ruling?

By the way, if the prize was going to be divided between the A and B players with 3.5, all 5 players ought to get the same amount.

I personally would say that the X, A and 3rd prizes go to the 4.0s and that the B prize is up for grabs between the 3 B players only. It makes sense to think of the 3 4.0s as one group and bring in the top 3 prizes. Then you don’t have to make the artificial decision as to which of the three brought in third place.

This is how it clearly would have worked if the prizes were $1 different, and it passes the equity test just fine to me. 3.5 for me was an expected score. 3.5 for a B player is more of an accomplishment in this field.

I see zero sense in giving the 3.0 expert anything. There is no logical basis to decide that it was the expert prize that was unclaimed.

1833

10

The final decision was that the 4.0’s will share the 3rd/X/A prizes, and that the 3.5 B’s will share the B prize. I now have to track down one of the players from North Carolina to get his share to him.

The really ironic thing is that if his friend (the other NC player who went 4.0 in the A section) hadn’t raised this issue Saturday night, I would have added the prizes in this exact manner; it was only the fact that I had been asked and then tried to ensure that I was doing the right thing by checking my logic here that caused all of this.

Lesson learned: Even NTD’s disagree on the proper course of action. :open_mouth: