Herb style qualifying = Funny math?

A picture of some weird statue dude isn’t going to help Fred qualify :slight_smile:

5 Likes

The heavily bonus-pointed systems are more vulnerable in theory to skewing towards someone with a few good results plus some trash, but in practice that rarely happens. Looking at numerous events from the few years, you just don’t see a player with 2#1’s, a #2 and 2 trash games qualifying in place of someone with all #3-#10 results. People who can get more than one first or second just don’t tend to be people that have trashy 4th- or 5th-best games. It can happen, and it’s likelier at PAPA than elsewhere, but in the ones I’ve looked it, it doesn’t. Maybe someone can find one out there somewhere, but I’d bet they’re pretty rare.

Does a picture of getting nuked help? http://tiltforums.com/uploads/default/219/c6678d324b6866ba.JPG

agreed, Bob. But look at Timothy Street in Karl’s modified Atlanta standings… His best game in his resume was a #3, and all of his other games were solid, but he misses out on any benefit of his #3 score vs. a #2 score in the current HERB scoring, but if you apply a standard .88, .9, .92, etc., Timothy gets in - so this isn’t about “riggin” the numbers to get someone across the cutline, it’s about fairness to someone’s resume that is just about the same as someone else both having an equal and fair chance to make the cut… And none of this, btw, is suggesting we take scores or their ratios into account - it’s all about your finishing place related to everyone else’s finishing place on that game.

I think the bar-napkin math analysis (which can get one into trouble when scribbling numbers down without detailed analysis of why something is happening) and what prompted Chris’ post was that if a player has an average score above 16th place on all games, then it is reasonable to assume one should finish in the top 16.

So 2 “trashy scores” should still theoretically work with (2) #1 scores and (1) #2 score if the scores all averaged in the top 16. Under the use 5 count 5 games system, 16th place is 72 points, so 72 x 5 = 360 points is a good guess (using the bar napkin math) as to what the cutoff should be. (2) #1 scores and (1) #2 gives 290 points toward the 360, so getting 70 more points across 2 games in a HERB style should be pretty easy for a player who is capable of three awesome scores.

Once the amount of games increases where the venn diagram intersection of games between players potentially separates, throw the bar napkin away.

As we already figured out, the cutoff at Pinvasion was 396 which is 79.2 points or averaging between 8th and 9th, so the typical rules of thumb and rationale don’t apply.

1 Like

Except we did figure out that since there were 9 games in use, not 5, that you have to average better than 16th since there are in essence 1.8 “sets” of 5 machines in play. You have to be in the top 8 on one set or the other, which makes perfect sense. In general, if you take N players for the playoffs and use best Y of Z machines, you “should” need to average NY/Zth place to get in, or in this case, 165/9 = 80/9 = 8.89th place.

I agree about the sets; hence my statement of “playing in parallel” earlier, or 8 players using one set and 8 another.

Afterwards the math is easier (and fun) to analyze, but when in the heat of qualifying, rules of thumb are what players are looking for to make decisions (unless you are like Keith or Bowen and 1 ticket and done). Most of the time, the quick estimate thought process during qualifying will be: If all of my individual scores are in the top N scores, I should be in the top N. It is a reasonable assumption-but the details always get in the way as we have seen.

Further, the rule of thumb NY/Z formula would not be accurate either at NDISC N=24, Y=6, Z =12 so, 246/12 =12 (which would seem to make sense using what happened at Pinvasion where the average was 8-9). The NDISC data shows 18th place average was sufficient. We also know that the averaging 24th per game was not sufficient. So on one hand, the first effort at bar napkin math would suggest too loosely and incorrectly that averaging top 24 is good enough and refinements of the bar napkin math show too strictly that top 12 is required.

This is a fun topic :slight_smile:

Jeff, once you brought up the 5 of 9 vs. a 5 of 5 it made a lot more sense to me. In almost all of Karl’s alternate scoring examples I fall short by one or two places. It truly was a case of just “play better”. Ironically, it was my own machine, Tommy, that kept me out.:flag_gb: During the last 90 minutes I had concluded that I just needed a score of 412M on Tommy to get into the top 16. I tried several times and had a really good game that was just 408M! Alas, I think I wasted too much time trying to better my score on Genesis, just because it consistently had the shortest line.:mens:

2 Likes

For the Rocky Mountain Pinball Showdown last weekend, I did a limited Herb, 20 attempts on an 8 game bank, taking the best 5 machines.

http://neverdrains.com/rmps2016/playerIndex.php

It was originally published that we would do 100-90-85…but switched to 100,97,95,93,92 a few days before the event. I think the huge bonus for first and second doesn’t make sense at all for limited events (though I’m not sure which type Pinvasion was).

My sense also is that it gives more hope to the lower/middle tier players who aren’t going to be able to pull off a 1st or 2nd no matter how hard they try…but I have no anecdotal data to back that up.

I want to see a scoring system where scores bleed up. Say in Atlanta you have 132 players, but not everyone plays every game. Lowest score on that game gets 1 point. If everyone plays walking dead, the top score for walking dead gets 132 points. You’re basically getting a point for every player each of your scores is better than for each individual game. Count all games involved, that might get interesting.

I’m sure this version has many flaws and it’s probably been done already. Still I would like to see what results would look like if done in this fashion. Thoughts?

1 Like

Dude, yes. The New Super Smash Bros. did a similar system to determine your ranking. It feels pretty great and everyone tends to feel like a winner. “I’m better than x players!” No matter the number, it feels better to think of your ranking positively as opposed to “this is how many people are better than me.”

1 Like

We did a standard divisor for the CAX tournament for several years. The prevailing opinion was negative, because you didn’t have a clear picture of what would happen to you if you got “bumped”: one bump could be no points at all while another could be a lot of points. Having all the “bumps” worth 1 point is a value that, based on the experiences we had, supercedes the accuracy of the exponential decay formulas.

Sorry, but this is a terrible rule for the “5 to make 9” format. If all players must play every game this format is identical to a normal 1-point ladder. But in the limited-entry format, unpopular games would end up completely unplayable, because first place on an unpopular game would be equal to 50th or worse on a popular game.

I would be perfectly fine with a decay of a single point for the entire range: 100,99,98,… But to arbitrarily assign large bonuses for the first two scores in that sequence seems super arbitrary to me.

Agree with Bowen 100%. In the case of Atlanta, here are the changes, just for reference. But it would affect who plays what game how much and completely muck things up. Goes in the “do not use” file for best Y of Z games.

Orig Rank Player Points Adj Rank
1 Trent Augenstein 405 5
2 Shannon Stafford 389 15
3 Justin Campbell 414 1
4 Jason Werdrick 390 13
5 David Riel 410 2
6 Bob Matthews 399 7
7 Joe Geneau 402 6
8 Nick Zendejas 407 3
9 Aaron Metz 389 15
10 Will McKinney 397 10
11 Steven Bowden 406 4
12 Derek Miazga 398 8
13 Derek Fugate 393 12
14 Lee Moscaritolo 389 15
15 Chris Compton 380 19
15 Sanjay Shah 382 18
17 Frederick Richardson 398 8
18 Chris Warren 390 13
19 Daniel Coyle 394 11

So I guess my response to that is, why should a player get more points on a less popular game? At pinvasion, top score of TWD would receive 102 points while the top score of genesis only receives 73. I want to see what competitor’s decisions would be in this scenerio as to what games to play and what games to ignore. I get the feeling that if pinvasion had this format, there would’ve been similar number of plays on each machine. I’d have to see an example of it failing before I dismiss the idea completely.

I know if i needed two more points and I had the top score on something, I would just go find two more people on the show floor and pay for them to play the game I had the top score on :wink:

3 Likes

Come on though. You wouldn’t do that, and neither would anyone else that isn’t a total jerk. And you could just file that under the rules that if caught you could be banned from the event or future events. I think the possibility of losing tournament playing privileges is deterrent enough. I like @PinballNarcissist 's idea a lot, but I have less experience than most on here.

Caught? What kind of rule could catch this? There isn’t one. Trust me, this would definitely happen, and there’s no proof or enforcement possible.

The ability to get top 10 out of 80 is not that different from the ability to get top 10 out of 120. It’s far different than the ability to get top 50 out of 120, which this proposal would make equal.