32B3 -- the Stupidest Rule in the Rulebook

Running Chess Tournaments
1

In the topic Rule 32B3 question, Steve Immitt sets forth the following prize distribution problem:

Overall prizes:
1st $2375
2nd $1149
3rd $536
4th $306
5th $230

Under-2400 prizes:
1st $1379
2nd $766

(The prizes could be a little less Insane than these, but the problem wouldn’t go away. We’ll stick with the above as a valid example.)

The top eleven players finished as follows:

5.0

4.5

4.0
4.0 (under 2400)
4.0 (under 2400)
4.0 (under 2400)

3.5 (under 2400)
3.5 (under 2400)
3.5 (under 2400)
3.5 (under 2400)
3.5 (under 2400)

How to distribute the prizes? The first two are easy:

5.0 wins $2375 (clear first)
4.5 wins $1149 (clear second)

What’s next? Under rule 32B3, the four players tied at 4.0 combine and split the highest four remaining prizes. These prizes include some of the under-2400 money, since some players in the group are under 2400:

Each 4.0 wins $747 (one-quarter of $536+$306+$1379+$766)

The remaining prize is split among the five players tied at 3.5:

Each 3.5 wins $46 (one-fifth of $230)

What’s wrong with this picture? If the three under-2400 players at 4.0 had each scored just 3.5, then the higher-rated 4.0 would be eligible only for the $536 third prize. He would have won less by outperforming three under-2400 players than he did by tying with them!

I call this an anti-monotonicity. Roughly speaking, a function is monotonic if it goes up when it looks as though it should go up, and down when it looks as though it should go down. Here, just the opposite happens.

Rule 32B3 is also replete with discontinuities – situations where a small change in the prize structure results in a huge change in the prize(s) won by one or more individual(s). These discontinuities have been documented in other threads.

So, if 32B3 is severely deficient in the continuous and monotonic departments, how does it fare in the awarding-class-prizes-to-class-players department? Well, let’s see.

A couple of posters in that other thread referred to “blue currency” and “pink currency”. I prefer a simpler, far more logical, approach. To figure out where the class prizes went, simply calculate all the prizes two ways: (a) as if there were no class prizes, and (b) with the class prizes included. The difference, player by player, is the amount of class money that went to that player.

Column A is each player’s prize, if there had been no U2400 prizes. Column B is each player’s prize, including the U2400 prizes:

[code]
Player: A: B: Difference:

5.0 2375 2375 0
4.5 1149 1149 0
4.0 268 747 479
4.0 (under 2400) 268 747 479
4.0 (under 2400) 268 747 479
4.0 (under 2400) 268 747 479
3.5 (under 2400) 0 46 46
3.5 (under 2400) 0 46 46
3.5 (under 2400) 0 46 46
3.5 (under 2400) 0 46 46
3.5 (under 2400) 0 46 46[/code]
Thus, $479 of under-2400 money went to a player rated 2400+.

What went wrong? Here is the wording of 25B3, the Stupidest Rule in the Rulebook:

If winners of class prizes tie with winners of place prizes, all the cash prizes involved should be summed and divided equally among the tied winners …

This wording has been around for a long time. The above version comes from the 1st edition (Morrison, 1975) of USCF’s Official Rules of Chess. It is repeated verbatim in the 2nd (Morrison, 1978), 3rd (Redman, 1987), and 4th (Goichberg, 1993) editions. In the 5th (Just) edition, the concept is generalized to include class winners in multiple classes:

If winners of different prizes tie with each other, all the cash prizes involved shall be summed and divided equally among the tied winners …

Now, pardon me, but isn’t there a conflict of philosophy here?

The whole idea of class prizes is that a class player can win more money for the same score than a player outside the class. For example, a B player in a single-section five-round event might well win the under-1800 prize with a score of 3-2, whereas a master or expert with that score will probably go home empty-handed.

You can claim that class prizes are unfair if you want to, and you will have a valid point. But please make that point elsewhere, such as in a new thread. Let’s keep this thread focused on the best way to implement the idea of class prizes, which is to allow a class player to win more than a non-class player for the same score.

It makes no sense, then, that “all the cash prizes … be summed and divided equally” among class and non-class players alike. This goes against the whole grain of class prizes.

Here is a better way:

A. First award the place prizes, without regard to the class or “under” prizes. In other words, figure out what the place prizes would be if there were no class prizes.

B. Next, award the class or “under” prizes to the players in those classes. If some players already receiving a prize in step A would now have their prize increased, get the increase from the class prizes, rather than tinkering with the place prizes already calculated. Observe the limit-one-prize-per-player rule, awarding any leftover amount to other player(s) in the class.

C. If there are class or “under” prizes in multiple classes, award prizes in the more inclusive categories before those in the less inclusive categories. For example, figure the under-2200 prizes before the under-2000 prizes.

How would this work in our Insane example at the top of this post? First, award the place prizes:

5.0 2375 4.5 1149 4.0 268 4.0 (under 2400) 268 4.0 (under 2400) 268 4.0 (under 2400) 268 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0
Now throw in the U2400 prizes:

5.0 2375 4.5 1149 4.0 268 4.0 (under 2400) 268 + 715 4.0 (under 2400) 268 + 715 4.0 (under 2400) 268 + 715 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0 3.5 (under 2400) 0
(Each 4.0 under 2400 gets one-third of the $2145 under-2400 prize fund.)

But this violates the limit-one-prize-per-player rule. Each of these three players is entitled to at most $268 or $715, whichever is larger, i.e. $715. The excess $268 is thus removed from each of these three players’ winnings. BUT we don’t touch the already-awarded place prizes. Instead, the excess $804 ($268 times 3) comes from the under-2400 fund, and is then redistributed to the five under-2400 players tied at 3.5. Each gets one-fifth of $804, or $161:

5.0 2375 4.5 1149 4.0 268 4.0 (under 2400) 268 + 715 - 268 4.0 (under 2400) 268 + 715 - 268 4.0 (under 2400) 268 + 715 - 268 3.5 (under 2400) 0 + 161 3.5 (under 2400) 0 + 161 3.5 (under 2400) 0 + 161 3.5 (under 2400) 0 + 161 3.5 (under 2400) 0 + 161
Voila! All prizes have been awarded, and 100% of the under-2400 prize fund has gone to the players under 2400.

And, gone are the discontinuities, the anti-monotonicities, and the general stupidity of the Stupidest Rule in the Rulebook.

Bill Smythe

2

So Bill, if only one player U2400 had finished with 3.5 pts then they would have got $804, more than everyone else but 1st and 2nd place?

3

IIRC, these rules evolved in the days when generally place prizes exceeded class prizes, such that a tie for 3rd without including class prizes would have diminished the resulting 3rd place prize payable to all players. Thus pulling the class prizes in was a way to minimize the diminishment by players who qualified for a second prize.

This solution can yield unintended results if the class prize exceeds the place prize.

However, I think your solution also yields unintended results. In essence what you’ve done can be viewed as you’ve taken the tied portion of the place prize from the player who could only qualify for one prize, and you’ve given it to players who didn’t earn that prize at all.

I think a better rule would be to simply say that the class prizes are added in and all are divided evenly, but to also include this principle:

That no player in the pool may win more than what they would have received if they had qualified either solely or in a tie for the largest prize for wish they were eligible.

For the dollars remaining in the prize pool, either of the following methods may be used and they arrive at the same result:

(i) Any excess above that limit is then divided evenly by the remaining players in the tied pool, with respect to the prizes for which they are eligible, and is added to the amounts determined in the initial tied pool prize distribution calculation.

(ii) The remaining prize pool (after allowing for any players and dollars removed from the pool due to the maximum limit above) is divided equally among the remaining eligible players.

In this case, that is equivalent to saying that the single player who earned 3rd would get it – 3rd would get the earned $536. The excess is $747-536=211, 211/3= $70 (and change) and the three ties would receive $817 each.

This is also equal to the sum of 4th plus the two class prizes divided by 3: 4th $306, 1st $1379 2nd $766 – and this is the other way to view it, that a player in the tied group received his full prize, so now he is eliminated from the tied group and the remainder of the group splits the tied prizes. In this way, the player receives the most he could, and the other tied players receive the most they could, and the class prizes are rewarded ONLY to the class.

In this way 3rd has no complaint - he/she is awarded his/her full advertised prize. The remaining 3 players split the prizes for which they qualify, and all the prizes go to the “appropriate section”. And the one prize per player rule is also maintained.

4

Bill, your example has the 2400+ 4.0 getting 268. If he had scored 3.5 (and nobody else scored as well) then the U2400s would split 3rd, top U2400 and second U2400. In that case the 2400+ would have gotten 306. Your method throws in another discontinuity where the 2400+ gets LESS if he does better.

5

Personally I’d be fine with this rule change/clarification (which matches what was done in the tournament).

It would be a clarification if the 32B3 phrase “for which others in the tie are ineligible” implies that the 2400+ is not eligible for 4th if he already was awarded 3rd.

6

Assume Bill’s situation with one change - a 5th person with 4.0 points who qualifies for place prizes only, and one fewer persons with 3.5

Overall prizes:
1st $2375
2nd $1149
3rd $536
4th $306
5th $230

Under-2400 prizes:
1st $1379
2nd $766
The top eleven players finished as follows:
5.0

4.5

4.0
4.0
4.0 (under 2400)
4.0 (under 2400)
4.0 (under 2400)

3.5 (under 2400)
3.5 (under 2400)
3.5 (under 2400)
3.5 (under 2400)

Now 5 people tie for 3-5 and the top two under 2400 prizes. In this case two players qualify for no other prize than a place prize. The largest that they would qualify for is (3rd $536 + 4th $306)/2 = $421.

The total pool is 3rd $536 + 4th $306 + 5th $230 + 1st U2400 $1379 + 2nd U2400 $766 = $3,217. Dividing by 5 tied gives: $643.40 So in this case the two place only qualifiers would receive $421. The remaining 3 would receive $791.67. Again, all class prizes have been rewarded to a class.

This is calculated by either:

  1. $643.40 - $421.00 = $222.40, then times 2 players receiving it = $444.80 divided by 3 players receiving = $148.27 adding to the $643.40 “average” = $791.67

  2. 5th $230 + 1st U2400 $1379 + 2nd U2400 $766 = $2,375, then ÷ 3 = $791.67

7

Hi,

I am not currently a delegate although I would be willing to be one for IL and go to the Open in Madison. However, in case I am not, is someone willing to sponsor or cosponsor the following ADM:

Prize Limitation In Ties to Ensure that Prizes Designated as Class Prizes are Paid Only to Eligible Class Players:

No player who is eligible only for a single prize type (for example overall prizes), but who is tied with a pool of players for multiple prize types (for example, overall and class prizes) may win more than what they would have received if they had qualified either solely or in a tie for the largest prize for wish they were eligible.

To distribute the dollars in the prize pool, either of the following methods may be used to determine prizes and they arrive at the same result:

Method A:

  1. Determine the prize due to each player in the pool, without the limit defined above.
  2. Apply the limit to the affected players.
  3. Any excess above that limit is then divided evenly by the remaining players in the tied pool who do not have the limit and is added to the amounts determined in the initial tied pool prize distribution calculation.

Method B:

  1. Determine the prize due to each player in the pool, without the limit defined above.
  2. Apply the limit to the affected players.
  3. Subtract the total aggregate prizes for the affected players from the original prize pool.
  4. Divide the new, net prize pool by the number of players who were unaffected by a limit to determine the prize for those players.

Example I:
Overall prizes:
1st $2375
2nd $1149
3rd $536
4th $306
5th $230

Under-2400 prizes:
1st $1379
2nd $766

The top seven players finished as follows:
5.0

4.5

4.0
4.0 (under 2400)
4.0 (under 2400)
4.0 (under 2400)

3.5

Method A:
1st receives $2,375
2nd receives $1,149

The ties are calculated as follows. 4 people tie for prizes, so to follow the 1 prize per player rule, 3rd, 4th, and 1st and 2nd under 2400 are added together: $536 + $306 + $1379 + $766 = $2,987. Divided by the four players is $746.75. One player qualifies ONLY for an overall prize, so this person is limited to $536. $746.75 - $536 = 210.75. This amount is evenly provided to the other 3 tied players: =$70.25. Added to their original $746.75, these three players receive $817 each.

Method 2
1st receives $2,375
2nd receives $1,149
One player is limited and cannot share in any portion of the class prizes. This person will therefore receive the $536 3rd place prize. The remaining ties players receive the two under 2400 prizes and 4th place: $306 + $1379 + 766 = $2,451. Divided by 3 = $817.

Example 2: Assume the same situation with one change - a 5th person with 4.0 points who qualifies for place prizes only, and one fewer person with 3.5

Overall prizes:
1st $2375
2nd $1149
3rd $536
4th $306
5th $230

Under-2400 prizes:
1st $1379
2nd $766

The top seven players finished as follows:
5.0

4.5

4.0
4.0
4.0 (under 2400)
4.0 (under 2400)
4.0 (under 2400)

Now 5 people tie for 3-5 and the top two under 2400 prizes.

Method 1
1st receives $2,375
2nd receives $1,149
The total pool is 3rd $536 + 4th $306 + 5th $230 + 1st U2400 $1379 + 2nd U2400 $766 = $3,217. Dividing by 5 tied gives: $643.40 So in this case the two overall-place-only prize-winners would receive $421 because (3rd $536 + 4th $306)/2 = $421. The excess that the overall players did not receive is $643.40 - $421.00 = $222.40 each. Multiply times 2 players receiving it = $444.80 divided by 3 players receiving = $148.27. Add this to the $643.40 “average” = $791.67

The remaining 3 would receive $791.67. Again, all class prizes have been rewarded to a class.

Method 2
1st receives $2,375
2nd receives $1,149
In this case two players qualify for no other prize than an overall prize. The largest that they would qualify for is (3rd $536 + 4th $306)/2 = $421. So these two players receive this amount.

The remaining three players split the under 2400 prizes, and 5th place: 5th $230 + 1st U2400 $1379 + 2nd U2400 $766 = $2,375, then ÷ 3 = $791.67

8

Kevin and Jeff have given me a lot to chew on here. Give me a few days.

Bill Smythe

9

I always thought that the point of class prizes were to provide a way that those who may not have hope of winning a prize otherwise might nevertheless be able to walk away with something. It’s an equalizer for those of us who will never be masters to nevertheless be encouraged to come out and play in tournaments.

The distinction is not that a class player goes home with more or less money than a non-class player for the same score, but rather that a class player might go home with money when that player would have otherwise went home with nothing. And players not eligible for a class should have sufficient strength to be in competition for place prizes.

Though I do agree that I have never liked that over-players could be awarded under-prize money.

I’m not sure why a complete separation wouldn’t work (is that what you meant by blue and pink currency? Couldn’t find that in the referenced thread.) A complete separation meaning place prizes are calculated and awarded, then class prizes are calculated and awarded separately - and it does not matter what was already awarded.

Get rid of the notion of the one-prize-per-player maximum, and if a class player ends up getting both a place prize and a class prize that’s a bonus reward for outperforming the field.

This would be the example above where the under 2400 4.0s walk away with $983 and the 3.5’s got $0.

Yes, in this case the under 2400s who scored 4.0 walk away with much more than the other 3rd place player. But I fail to see the problem with that idea - the top players got the award based on how they played, and the class players shared in the reward for being the best in their class down to the prize fund limit. The 3.5 players got nothing, but welcome to the playing world that 90% of us inhabit.

In practice, I’m sure such rules would alter the way we calculate and offer prize funds for classes.

But, in fact, I don’t have a problem with the solution as it was originally presented. At least there the best players by performance got the most money and the 3.5 players walked away with something. Break the suggestion that class prizes should be less than place prizes, and you get oddness like this.

(If we’re playing ‘shoulds’, it should be acknowledged that the prize structure wasn’t in accord with the rules.)

Final note… rather than advocate ‘the best’ way, though (I think we’d all agree there is no such single solution for everything from club play to the U.S. Championship?,) can one not devise whichever system one wants to and then publicize heavily the variation? Then let the marketplace decide. (So I’m saying, Bill, flesh out your proposal as a series of alternate rules and TD Tips that an organizer can copy and reproduce as a published variation for a tournament. I think you’re practically there already.)

10

But this is a very common misconception. “Over players” are never awarded “under-prize money.”

The “blue currency” and the “pink currency” was my attempt to explain why this is a misconception. The idea was that the general prizes were paid in “blue currency,” while the “under prizes” were paid in “pink currency.” If it is not possible to pay the prizes with all the pink currency (and some blue currency, perhaps) going to the “under” players, then the “under players” should be excluded from the general prizes and just divide the “under prizes” among themselves.

Allow me to give two examples. For simplicity, let’s assume the prize fund is just two prizes: first place, $300; top under 2000, $200.

Case 1: There are two scores of 3.5, with one player rated 2100 and the other rated 1950. In that case, the prizes can be paid by giving the 1950 $200 of pink currency and $50 of blue currency, and giving the 2100 $250 in blue currency.

Case 2: There are three players scoring 3.5, rated 2100, 2050, and 1950. If we divide the sum of the two prizes evenly, then we would pay the 1950 player $166.67 in pink currency. That leaves $33.33 of pink currency to be combined with $300 of blue currency, distributed between the 2100 and 2050. (Someone needs to put in a penny, which I will assume is copper in color. :slight_smile:) In this case, the correct prize distribution is to award the 1950 the entire $200 prize (paid in pink currency) and to pay the 2100 and 2050 each $150 in blue currency.

Of course, this generalizes to the case that several different classes are involved in a tie. Each class is assigned a different color currency. If the “sum and divide equally” method produces any case of a color of currency going to players not eligible for that color, then the eligible players should be taken out of the general split and instead divide their color currency among themselves (since they will win more under that distribution).

11

I agree that they shouldn’t be, but…

In Smythe’s original example, some Blue players were paid some Pink currency. Bill came up with an idea that addressed this, but that also provided some “undeserving” Pink players Blue currency. I think the idea I came up with ensures that Pink currency only went to “deserving” Pink.

12

I have to confess that I have not had time to study Bill’s original example carefully. However, if the prize distribution ended up awarding pink currency (“under” prize money) to players not eligible for the under prizes, then the prize distribution was incorrect. The words “unless any of the winners would receive more money by winning or dividing only a particular prize for which others in the tie are ineligible” ensure this. (Of course, I’m extending “a particular prize” to read “particular prize(s)” to allow, for instance, two “under” prizes to be combined and divided among the “under” players.)

13

OK, I actually had looked at Bill’s original scenario in the thread started by Steve Immitt and commented on it extensively in that thread.

At first, it would seem that blind application of rule 32B3 (sum the prizes and divide evenly) would indeed award under-2400 prize money to the one player in the four-way tie not rated under 2400. (Bill, please check your original problem statement. I believe that the situation Steve Immitt described had four players, not five, tied with 4.0. Three of these players were rated under 2400.)

In fact, I originally stated my opinion (which was my own opinion as an individual, not a formal opinion of the rules committee) that the sum-and-divide was correct based on what seems to be very clear wording of rule 32B3. However, I was given pause by the observation that the player who was rated above 2400 would win more prize money because of the tie than if that player had been the only one with a score of 4.0. Indeed, it does seem nonsensical that a tie should increase a player’s prize.

I put essentially this question (but with the prize amounts changed to round numbers to make the calculation easier) to the rules committee. I did not generate as much discussion as I had hoped to, but the nearly unanimous reaction was that sum-and-divide of all four prizes (3rd, 4th, and the two under-2400 prizes) was incorrect. Instead, the player rated over 2400 should receive the entirety of the third place prize, and then the three players under 2400 should split fourth and the two under-2400 prizes equally.

Notice that the sum-and-divide of all four prizes does not violate the “blue/pink” constraint! It is possible to pay the over-2400 player entirely in blue currency by paying the third place prize and a portion of the fourth place prize. Then, the three under-2400 players each receive one-third of the pink currency plus the remainder of the blue currency divided equally three ways. (Hence, there is no violation of the “unless …” clause of 32B3.)

The problem is that the prize payout to the player rated over 2400 (which must be done entirely in blue currency) violates 32B1. By receiving all of the third place prize plus a fraction of the fourth place prize, that player is winning more than one prize.

While this situation does seem to be covered by a combination of rules 32B1 and 32B3, I do think 32B3 would be well served by adding language such as “provided no player wins more than he would receive if there were no other players tied with him.” (Please see Steve Immitt’s thread for the exact wording, including a good suggestion for improvement by Jeff Wiewel.) I intend to submit that to the rules committee for discussion, with the goal of making it an ADM for 2013.

14

Then you may want to take a look at my wording above, which is refined from the earlier thread, but which may still need additional refinement; for example, your concept above is still too general - there could be other “overall” players tied with him as well as other “class” players tied with him.

I’m also not sure that my wording has fully taken into account what might happen if we have the “more normal” (or at least more intended/traditional) idea of the class prizes being less than the overall - in which case this stuff doesn’t happen. We don’t want to inadvertently impact what we are already doing in THAT situation. I’m hoping that the wording doesn’t require that we take it piece by piece.

15

It seems to me we’re saying essentially the same thing. If a class player wins money he would not have won if he weren’t in that class, that means he is winning more ($X, where X is non-zero), than he would have ($0) if he were not in the class. When I say that the class player wins more money than a non-class player with the same score, I am simply generalizing to the case where the second amount in question is not $0.

It’s in this post by Ken Ballou.

If I’m not mistaken, that’s what Kevin is proposing also. In other words, get rid of the “limit one prize per player” rule.

And that just might be a good idea. From some of the other posts here, it now seems that retaining the “one prize” rule can produce the same sort of discontinuities, anti-monotonicities, and just plain illogic that already plagues the Stupidest Rule in the Rulebook.

On the other hand, I have a certain sympathy for the “one prize” concept. It spreads the prizes around a bit more – even if it sometimes awards prizes to players who may seem undeserving.

Amen.

I don’t see why that prize structure was out of compliance – unless you’re saying it’s against the rules for the top class prize to be higher than the bottom place prize, but I’m sure there’s no such rule.

Of course. I’m hoping this system, or some modification thereof, will eventually appear in the rulebook, at least as a variation with a specific name. Poster Jedi coined the phrase Cascade System, which I think is perfect – thanks, Jedi. All an organizer would need to do is say “This event will use the Cascade System of prize distribution”, perhaps with a link to one of these threads.

Yes, we’re pretty close. But some of the bugs that have been pointed out may require a brief return to the drawing board.

OK, but the most logical way to figure out which currency is blue and which is pink would be simply to calculate the prizes two ways, with and without the class prizes. The difference, for each player, is that player’s pink currency. In the original example, the 4.0 player over 2400 was winning pink currency, which is my fundamental objection to the Stupidest Rule in the Rulebook to begin with.

Check again. The two presentations are the same. In both, there were four players tied with 4.0 and five with 3.5.

IMHO, the Stupidest Rule in the Rulebook is so bad that it must be completely replaced, not just tweaked. The proposed Cascade system should work just fine, provided the problems with the “limit one prize per player” concept can be worked out. Perhaps one solution would be to re-define the maximum that any player can win. The maximum could be defined, for example, as the greatest single prize the player is entitled to, without regard to ties. (I think this was one of Kevin’s ideas, I’ll have to go back and throughly review that post.)

So, it’s back to the drawing board for now. Stay tuned.

Bill Smythe

16

I think in the above posts, “one prize per player” is being misunderstood. One prize per player means that one prize per player goes into the pool. The dollars a player receives might be from more than one prize - i.e. they may have (partially) won more than one prize.

My approach still applies the one prize per player rule.

17

I believe that there needs to be more substance to the idea behind the proposed motion presented here. One concern is that the “law of unintended consequences” will kick in and muddy the waters. I have observed that when a major change is in the air for a USCF rule change/implementation it takes at least two to three years to flesh it out after it’s initial passage; i.e., the law of unintended consequences kicks in and needs to be addressed (hey, look at the how long it took 14H to settle down to it’s present form or the “write the move after making it” debacle).

Next, to sell this philosophy of prize fund distribution to the Delegates the motion maker needs to remember that many Delegates will vote based on their understanding of the new improved rule and “how it impacts them and those they represent.” They will have a lot of questions (technical and mathematical) that will need to be addressed.

One technique to consider here is to get any motion based on this philosophy installed as an “announced” variation. The military team version of “tiebreak systems” that passed a year or two ago comes to mind as an example.

18

Tim, I agree with everything you said, especially about the law of unintended consequences.

In fact, the present version of 32B3 (in effect since at least 1974) is a prime example. It has caused more unintended consequences than probably everything else in the rulebook combined. This forum is littered with lengthy, argumentative threads about the problems it has caused.

I also agree that, if anything like the Cascade System is to be implemented any time soon, it must start out as a variation to be announced in pre-event publicity. Then, after a year or two, it can maybe become the mainline rule, depending how well it works out. Or maybe not. Or maybe it will require still further modifications.

One must, however, avoid the temptation to shy away from new ideas just because they are new ideas and may have a few unintended consequences. The Stupidest Rule in the Rulebook is a disaster, albeit a 40-year-old disaster. It sorely needs to be changed.

Bill Smythe

19

I guess I would add that the rule I proposed (that a player cannot win more than the advertised prize for which he/she is eligible) was proposed specifically to eliminate an unintended consequence of the current rule. If we test every rule change for a couple of years before implementing it, we’ll never make a change.

In programming (machine rule development) one tests a program by creating all the possible situations and testing them. That eliminates unintended consequences.

I’m not saying that the proposed rule has yet gone through that rigor; but this is precisely why we need to take a new and simpler approach to rule development.

20

Good goal.
When programs get complex enough that the possible combinations reach into the millions (or higher orders of magnitude) it becomes difficult. Then you start grouping options together and hoping that your groups are valid.