Extra prize money after limit on unrated player's winnings

Running Chess Tournaments
1

I just saw this situation and did not find a clear statement in the rules to guide me.

Suppose an “under X” section of a tournament has guaranteed place prizes of $120-$80-$40. Suppose there is a two-way tie for first place. Of course, normally the two players would each win $100.

Now, it happened that one of the players was unrated, and the club has a policy that an unrated player can not win more than $50 (except in the open section).

How should this be handled?

  • Ken Ballou
    MetroWest Chess Club
2

Why would someone design a tournament prize payout that just BEGS to cause trouble??

3

Second. I can see four options, none especially attractive.

  1. Give the Unrated $50 and the other tied player $150. This means the latter is getting more than clear first, which doesn’t sound right.
  2. Unrated $50, other tied player $120, keep the rest. No comment needed.
  3. Unrated $50, other tied player $120, divide the remaining $30 proportionately among all the other prizes. Equitable but imposisbly complicated.
  4. Unrated $50, other tied player $120, pass on $30 to the next player(s) in line (so third is $70 instead of $40). Probably the best solution here. Note, however, that it would be possible to construct a prize fund in such a way that this would give the third-place finisher more money than the guy who tied for first, which can’t possibly be right.
4

I don’t know, but one sees stuff like this all the time. A lot of people simply aren’t “wonk-ish” enough to figure out the details of what they’re getting into when they design prize funds.

In the stated example, a better policy would be to have a separate Unrated prize in the Reserve section, along with a statement that “Unrated players are eligible only for place prizes in the Open or the unrated prize in the Reserve”.

Bill Smythe

5

Well, this is actually an exceptional case. The tournament in question had three sections: open, under 1700, and under 1400. The section in question was actually the “under 1700.” The unrated player won his first four games and had a zero-point bye in the fifth (final) round. Had he played the fifth round, it’s entirely possible he would have had a clear first place result with no tie.

It’s not every day one sees an unrated player with a Class A performance rating!

6

I think we should introduce a fith option to the solution:

Give the unrated $50 and consider him out of the prizes (meaning that you didn’t give him a first or second place prize). Since your overall prize fund WAS $240 for that section, take the $50 out of it, and give first place $95 (more than he would have gotten if the unrated guy really was rated) second place (although it is really third place) $63 (more than he would have gotten if he got a third place prize) and give third place (really fourth place) $32 (more than he would have gotten since fourth place really doesn’t have any prize). This way, the unrated gets the most he can get, and everyone else got more than they would have gotten if the unrated was rated.

Just my inexperienced opinion, but I think this is a pretty good solution to the problem at hand

7

Chesswolf,

I’m not quite sure I understand your reasoning.

Let’s suppose the unrated player was in fact rated. Then, I believe the correct procedure is to combine the first and second place prize money ($120 + $80 = $200) and divide that equally between the two players (so each wins $100). Then, the $40 is awarded to the third place winner. (In fact, there was a 3-way tie, so each would receive $13.33 [and the TD pockets $0.01 :slight_smile: ].

I don’t follow how you determined the rated player should win $95.

8

What he’s trying to do is subtract the $50 that goes to the Unrated, then divide the remaining money in the same proportion as the original prize fund. There are several problems with this, but the main one is that it results in the guy who tied for first getting less than [(1st + 2nd)/2], which is obviously wrong.

Your distribution is reasonable (though I think the player who tied for first should get $120 rather than $100, siince the third-place players(s) have no real claim on any of that 1st/2nd money). But, as I pointed out above, it would be possible to construct a prize fund such that doing it that way would lead to absurd consequences.

9

ah, yea, I did give the first place guy less than he would have gotten if the guy was rated. Hmm, and I thought I was good at math, lol. Anyway, yea, I agree, maybe the other $30 should go partly to the third place guy and partly to the fourth place, I donno, but I guess that is why I’m not a TD yet, lol.

10

The correct answer, equitable or not, is:

rfeditor option #4:

Unrated: $50
1st: $120
2nd-3rd: $70 or split among those tied.

The unrated, by the structure of the tournament, can win the 3rd place prize, but cannot win the 1st or 2nd (clear)
Very similar to what was said above. In essence the unrated is “eating up” the 3rd place prize (plus 10 bucks)
You then are basically awarding 2nd place to the next (rated) player (less the 10 bucks)

By the way “The club has a policy that…”
I hope this was advertised that way in the TLA for this particular event, or the unrated is due his full prize.

11

It’s not that simple. Suppose the prize fund had been $120 1st, $100 2nd, $90 3rd. Combining 1st-2nd and subtracting $50 leaves $170. Subtract $120 from that (the amount of clear first) for the other tied player, and you have $50 left. You cannot add $50 to third and give the third-place finisher $140. I suppose you could “invent” a 4th-place prize, but how much of that $50 stays with 3rd? $29? $1? None?

Of course, the solution is to think this through before designing the tournament.

12

its never that simple, you must follow some logic. It is always a simple thing to invent a structure that doesn’t work with a certain scenario.

That is not what I did in the example above. I subtracted the $50 from the bottom of the pool for the unrated, not the top. I left the top prize(s) intact, as much as possible.

The unrated cannot even win the 3rd place prize (clear) in this case. Take the $50 won from the 3rd place prize, and distribute the rest accordingly.
i.e. 1st, $120; 2nd $100; 3rd the remaining $40 (among the remaining players). if 2nd-3rd are tied, then split the $140.

Note that the 3rd place winner of $40 is really the 4th place in the tournament. If the unrated would have been pooled with the other two, that player would have got Zero. So he doesn’t have anything to complain about either.

I am sure that if you think hard enough (oh, about 30 seconds ought to do it) you can come up with something that doesn’t work here, either. But don’t bother. Like stated earlier in the thread, a reasonable organizer will avoid this type of dilemma

13

I agree with your result (at least in the original case), but not with the reasoning. The unrated tied for 1-2, so “his” $50 should come out of the 1st-2nd pool. Now, if the restriction had been phrased something like “Unrateds may not win more than 3rd place,” it would be a different matter. My point was that, while we can arrive at an acceptable answer by examination, there is no good rule (in the mathematical, not the chess, sense) for dealing with the general case.

14

I agree with David Kuhns’s solution. As I see it, the purpose of putting a restriction on prizes that can be won by an unrated player is to protect the rated players from unrated sharks who play in sections that are below their playing strength. Therefore, the prize for the unrated player should be calculated first and taken out of the prize pool, starting at the bottom, and then the remaining prizes should awarded as if the unrated player weren’t in the section. That way a player who ties for first with, or finishes second to, an unrated player will receive the same prize as if he were clear first.

As David said, in the original scenario the prizes should be: unrated $50, 1st $120, 2nd $70 (although I’d announce them as 1st-2nd $120/$50, 3rd $70). In the new scenario the prizes would be unrated $50, 1st $120, 2nd $100, 3rd $40 (announced as 1st-2nd $120/$50, 3rd $100 4th $40 - yes, I’ve done this, inventing a new prize).

In tournaments like the Massachusetts Open we say “Unrated players can play in any section, but may not win 1st or 2nd prize or title except in Open” (there are three prizes in each of the lower sections). This makes it simple, at least in this case: 1st $120, 2nd $80 (or $100), 3rd $40 (or $90) with the unrated player winning $40 (or $90).

In some tournaments that we (Massachusetts Chess Association) run there is a different method, which I don’t like: “Unrated players eligible for only 50% of prize except in Open”. This can make things tricky. In the original scenario ($120-80-40) if the unrated player tied for first I’d first calculate the prize that the unrated player would receive if he were rated: $100. Then take half of that amount: $50. Now award the remaining prizes as if the unrated player weren’t in the tournament: 1st $120 2nd $70 (announced as 1st-2nd $120/$50, 3rd $70). If the unrated player were clear first he’d win $60 and the other prizes would be 1st $120, 2nd $60 (announced as 1st $60, 2nd $120, 3rd $60).

Bob Messenger

15

I’ve thought of a problem with the “take the prizes from the bottom” algorithm for dealing with unrated prize restrictions. It works for the simple examples given earlier, but consider this one:

$$ 1200-600-300-100-100. Unrated can’t win more than $500.

Assume the top five players are A, B, C, D, and E, and A is unrated. The “take prizes from the bottom” algorithm produces this result:

A: $500 B: $1200: C: $600

We announced five prizes but we only awarded three. This can’t be right.

Instead I’d like to propose the following algorithm, which also allows for multiple unrated players and class prizes. Unfortunately it’s fairly complicated. It’s pretty much what I’ve been doing in all the tournaments that I’ve directed, but until now I hadn’t considered some of the more obscure cases.

Step 1: Calculate the prizes as if the unrated players weren’t in the tournament. This establishes the maximum prize for each rated player. In practice these prizes don’t have to calculated until they’re needed in step 2E.

Step 1A: If unrated players are limited to a percentage of prizes won, calculate the prizes as if all players were rated and then multiply the prize of each unrated player by the percentage. This establishes the maximum prize for each unrated player. Again, these prizes don’t need to be calculated until they’re needed in step 2D.

Step 2: For each score group starting with the highest one, until all unrated players have been awarded prizes and there is no more excess prize money:

Step 2A: Pull one prize for each tied player into the prize pool (or pull in a single prize if there is only one player in the score group).

Step 2B: Add in any excess prize money from the next higher score group.

Step 2C: Divide the prize pool equally among the tied players.

Step 2D: Apply prize restrictions for unrated players. If prizes are reduced take the unrated players out of the prize pool and go back to step 2C.

Step 2E: Apply prize restrictions for rated players. This may result in excess prize money which will go to the next lower score group.

Step 2F: If any players in the score group would win more money by taking a class prize, remove them from the prize pool and go back to step 2C.

Step 2G: Award prizes to any players left in the prize pool, and proceed to the next score group (step 2).

Step 3: If there is excess prize money after all prizes have been awarded, create a new prize and divide it among players in the next score group (step 2).

Step 4: Once prizes have been awarded to all unrated players and there is no more excess prize money, award the remaining prizes normally.

How does this work in practice? Let’s look at the example which I gave earlier:

$$ 1200-600-300-100-100. Unrated can’t win more than $500.

Assume the top five players are A, B, C, D, and E, and A is unrated.

I’ll show just the steps which are relevant.

Step 1: prize limits for rated players

B: $1200 C: $600 D: $300 E: $100 (F etc.: $100 - won’t be needed)

Step 2: top score group

Step 2A: $1200, only A is in the score group
Step 2D: A is unrated so he’s limited to $500. Excess is $700.

Step 2: second score group

Step 2A: $600, only B is in the score group
Step 2B: Add the $700 excess increases the prize to $1300
Step 2E: B’s maximum prize (step 1) is $1200. Excess is $100.

Step 2: third score group

Step 2A: $300, only C is in the score group
Step 2B: Add the $700 excess increases the prize to $400.
Step 2E: C’s maximum prize is $600, so he can win the entire $400.

Step 4: since there are no more unrated players and no more excess prize money, award the remaining prizes normally: D $100, E $100.

The result is:

A: $500 B: $1200 C: $400 D: $100 E: $100

Can anyone think of prize structure where this algorithm wouldn’t work?

Bob Messenger

16

The need for such a complicated algorithm speaks volumes to the inadvisability of allowing unrated players in lower sections to begin with.

In multi-section tournaments with significant prizes, the policy should be EITHER:

Unrated players must play in the top section or in the Unrated section (a separate section in which only unrated players are allowed to play).

OR:

Unrated players may win only place prizes in the Open section, or any specifically announced Unrated prizes in other sections.

What bothers me about this whole discussion is that nobody seems concerned about the unrated player’s last opponent or two. The poor guy who lost to the unrated in the last round was probably knocked out of the prize fund entirely, when he would have won that round (and a significant prize) if the unrated weren’t in the tournament. Maybe THAT’s where the remaining undistributed prize money should go.

Bill Smythe

17

For historical reasons which would take too long to describe, Massachusetts is big on many-many section tournaments. To say this is good or bad would be simplistic, but it has consequences.

If most of your tournaments are in six or seven sections, you can’t have an unrated prize in each section. So your choices are to limit the unrateds to one or two sections (typically top and bottom), or come up with some complicated formula to restrict prize money. The drawbacks to the latter have been made plain in this thread. The former solution is more common, but it will put some new players in sections clearly out of line with their strength.

Trying to find a “perfect” solution to this pointless. New unrated players are a temporary aberration to the system, and any rule flexible enough to deal with all of them would be too complicated to be useful.

18

Seems to me the best idea (as long as there are six or seven sections anyway) is to add an eighth section, just for unrated players, and require unrateds to play in that section or the top section. (If they play in the top section, make them eligible only for place prizes, not class or under- prizes.)

The entry fee in the unrated section can be small compared to the other sections, and there can be a small prize or two. Even if only a few players play in that section, the section can still be held.

Bill Smythe

19

Goichberg has tried this in a few tournaments, but for almost everyone else the section would be too small. (By “too small,” I mean “not enough players to make legal pairings.”)

20

It’s a tradeoff. On the one hand we want to encourage new players to enter our tournaments, but we don’t want strong unrated players to walk away with all the prize money. Yes, it’s hard luck on the guy who’s knocked out in the last round because he has to play a strong unrated player. That’s the risk that we’re taking. If we get enough negative feedback we’ll have to consider changing our policy.

Although the Massachusetts Open is the state championship it’s not really a big money tournament. Third prize in the lower sections (U2000 through U1500) is only $100 and the entry fee is $49, so there isn’t a big incentive for an unrated player to come in and “clean up”.

I agree that there need to be more severe restrictions on unrated players in big money tournaments, e.g. at Foxwoods unrated players could only play in the Unrated section or the Open. But for smaller events it makes sense to let new players play in a section that roughly corresponds with their strength. They have to get their ratings somewhere.

Bob Messenger