Herb style qualifying = Funny math?

Events
1

So I failed to qualify at the recent circuit event Pinvasion III. Due to perceived time constraints only the top 16 were advanced to the finals. I qualified 18th. Here is the funny math part. My five best games were a #4, #6, #11, #11, and #14. I would certainly think those 5 scores would get a player into the top 16, but it was not to be.

The scoring was the standard 100, 90, 85, 84, etc. Has competition gotten so tough in this format that you must have a #1 or a #2 just to get in?

Maybe Bowen could dissect the math on this one? It just seems odd to me.

2

Derek and Lee qualified with 5th place as their best finishing position on a game, and Sanjay had a couple of 4thā€™s as his best games in his run.

I think the ā€˜back endā€™ of your best tickets cost you a qualifying spot far more than getting a top 2 spot. Iā€™ll usually find that the top 8 eat those top spots up like candy, especially if there are byes on the line (which in PV3ā€™s case there wasnā€™t).

3

Good topic.

I think its more of the fact that tournaments are going towards best x of x (best 5 of 9 in Pinvasions case). Before it was always count them all. Now that the games are spread out where not everyone has to play/focus on all of them, it changes things a lot. Someone with better math skills can probably dissect it more but its my belief that it all stems from best x of x rather than all count. Its not that the format is inherently getting harder, its just changing with the integration of more games.

1 Like
4

Yes, this. You may think youā€™ll make top 16 from all top-16 scores, but thatā€™s not true when top 5 of 9 games get through.

For example, at CA Extreme last year, it took an 80-point average game (18th place) to qualify among the top 24 finalists. Qualifying at Pinvasion took an average 9th-place qualifying finish.

For example, Daniel Coyle finished with a 5th, 7th, 10th, 12th, 13th and didnā€™t qualify for Pinvasion.

Also the time constraints werenā€™t ā€œperceivedā€, they were correctly anticipated. The tournament with 16 finalists ran just within its time limits and Iā€™m glad that choice was made.

1 Like
5

My main ticket was 3rd, 15th, 16th, 17th, 18th. If my 3rd was a 1st I still would have needed ~10 more points to make finals. In a fake world where I can get those points and nobody elseā€™s points are affected, I still would haved needed 1st. 13th, 14th, 15th, 17th to get 397 pts.

Keeping the 3rd I could have done 3rd. 9th, 10th, 11th, 11th.

I needed to play better!

6

There were 9 games but only 5 are used and only the top 16 advance. With that many extra games in play it is almost possible to have 1 group of players qualifying on a set of games completely different than a second group. It is as if two groups are playing in parallel with little overlap.

The cutoff was 396 which is an average of 79-80 points per game which is a top 8 or 9 score required on every game. So it is almost like 8 players qualify using one set of games, 8 using another.

Perhaps 9 choose 6 or maybe 7 choose 5 works better. We have already figured out 5 choose 5 stinks with many players :slight_smile:

1 Like
7

This is why the points awarded to second/first should adjust based on the ratio of games counted:games available. The more games you have available and the less games you need to play, the lower the advantage for a first should be.

So for instance at CAX this year we are only taking top 5 of 12 games. This is great for keeping lines down, but if we did a traditional 100/90/85 it would mean that there would be way too much focus on getting a first on a game. So we did 100/97/95. A lot of other tournaments have been going this route as well.

7 Likes
8

I think another factor in CAX vs Pinvasion was a modified scoring 100-94-91-90ā€¦ In that setup, 9 points separates 1st and 3rd as opposed to 15 points in the 100-90-85-84ā€¦format.

Maybe someone can run their alternate universe scoring to see how things would have panned out at Pinvasion using CAX method, or alternatively what would have happened if CAX used 100, 90, 85ā€¦

(I think heyrocker and I were typing the same things at the same time :slight_smile: )

9

Did someone say alternate universe? :wink:

http://www.neverdrains.com/pv3/playerIndexAlt.php

8 Likes
10

I remember the first time a few years ago, when the first 5-out-of-9 HERB was tried in a circuit event (because 5-out-of-5 didnā€™t work anymore with the huge turnoutsā€¦) and watching the standings with @pinwizj, we realized how crazy it was that there were 9 sets of 5 and 10 point bonuses out there and they would totally screw somebody out qualifying, even if they put in 5 top 10 scoresā€¦ It was after that, that I always petitioned the HERB-style tournaments I was involved in, to neuter those bonus points down based on the ratio of games needed to games available to play. So for a 6-of-10, I would use 100-94-91-90ā€¦ and for 5-of-9, I would go down to 100-96-94-93.

The fact that there are any bonuses at all for the top two scores has always seemed silly and arbitrary to me as a player. PAPA has used that scoring for years, and itā€™s fine there, because their format is NOT HERB and those bonus points donā€™t really have much bearing on the outcome of qualifying, because most of those bonus points get lost on tickets that donā€™t count (for instance, of 120 scores of the top 24 players, at PAPA 19, there were only eight first or second place scores that affected qualifying, and at PAPA 18, there were only five). But for HERB, all of those points will count every tournament and it wasnā€™t even that bad when everyone had to play all the games, but when there are more games available it really skews the results. And not in a good way, IMO.

I think weā€™ve had plenty of threads here already that discuss using a simple curve function to decrease the points awarded for each finishing score in a standardized way and other than itā€™s ā€œcomplicatedā€, I donā€™t understand why its not used over the silly and arbitrary 100-90-85-84ā€¦ other than, you know, change is hard and itā€™s much easier to keep doing it the way ā€œitā€™s always been doneā€, even if thereā€™s no good reason for it.

Notice anything about what it takes to get a top eight bye at Louisville lately ? lol

http://www.neverdrains.com/lax2016/playerIndex.php

And we can now add Pinvasion III to that: notice anything about what it takes to make the cut in Atlanta lately ? lol

http://neverdrains.com/pv3

I guess I have to add this one to my list of crusadesā€¦

#ScoreHerbCorrectly!

3 Likes
11

#ScoreHerbCorrectly being that it actually doesnā€™t matter? :slight_smile:

Between the 3 models that Karl listed, Fred would have qualified in the other two versions (he didnā€™t in 100-90-85), with Chris Compton being the casualty in those situations.

All other 15 qualifiers would have been the same regardless of system.

I remember going through those numbers from our Gameworks results back in the day, and every point structure I threw at it resulted in that >90% ā€œsame people qualifiedā€ rate.

Karl - Can you throw up 100-99-98-97 just to take it to the extreme the other way?

Ultimately Iā€™m still of the camp that it just ā€œdoesnā€™t matterā€, because it doesnā€™t matter . . . unless youā€™re Fred . . . and then well youā€™re Fred . . . so thereā€™s a good chance that if the original qualifying scoring was 100-97-94 Fred would have found a way to only have qualified if 100-90-85 was the system used :stuck_out_tongue:

12

None of those alternate universes that Karl shows remove the ā€œarbitraryā€ top 2 scores get bonuses! Hey Karl, try it with one of these!! (Steady decay rates of 100 points with a .88, .90, .92, .94, and .96)

   1         100         100         100         100         100
   2          88          90          92          94          96
   3          77          81          85          88          92
   4          68          73          78          83          88
   5          60          66          72          78          85
   6          53          59          66          73          82
   7          46          53          61          69          78
   8          41          48          56          65          75
   9          36          43          51          61          72
  10          32          39          47          57          69
  11          28          35          43          54          66
  12          25          31          40          51          64
  13          22          28          37          48          61
  14          19          25          34          45          59
  15          17          23          31          42          56
  16          15          21          29          40          54
  17          13          19          26          37          52
  18          11          17          24          35          50
  19          10          15          22          33          48
  20           9          14          21          31          46
  21           8          12          19          29          44
  22           7          11          17          27          42
  23           6          10          16          26          41
  24           5           9          15          24          39
  25           5           8          14          23          38
  26           4           7          12          21          36
  27           4           6          11          20          35
  28           3           6          11          19          33
  29           3           5          10          18          32
  30           2           5           9          17          31
  31           2           4           8          16          29
  32           2           4           8          15          28
  33           2           3           7          14          27
  34           1           3           6          13          26
  35           1           3           6          12          25
  36           1           3           5          11          24
  37           1           2           5          11          23
  38           1           2           5          10          22
  39           1           2           4          10          21
  40           1           2           4           9          20
  41           1           1           4           8          20
  42           1           1           3           8          19
  43           0           1           3           7          18
  44           0           1           3           7          17
  45           0           1           3           7          17
  46           0           1           2           6          16
  47           0           1           2           6          15
  48           0           1           2           5          15
  49           0           1           2           5          14
  50           0           1           2           5          14
  51           0           1           2           5          13
  52           0           0           1           4          12
  53           0           0           1           4          12
  54           0           0           1           4          11
  55           0           0           1           4          11
  56           0           0           1           3          11
  57           0           0           1           3          10
  58           0           0           1           3          10
  59           0           0           1           3           9
  60           0           0           1           3           9
  61           0           0           1           2           9
  62           0           0           1           2           8
  63           0           0           1           2           8
  64           0           0           1           2           8
  65           0           0           0           2           7
  66           0           0           0           2           7
  67           0           0           0           2           7
  68           0           0           0           2           6
  69           0           0           0           1           6
  70           0           0           0           1           6
  71           0           0           0           1           6
  72           0           0           0           1           6
  73           0           0           0           1           5
  74           0           0           0           1           5
  75           0           0           0           1           5
  76           0           0           0           1           5
  77           0           0           0           1           4
  78           0           0           0           1           4
  79           0           0           0           1           4
  80           0           0           0           1           4
  81           0           0           0           1           4
  82           0           0           0           1           4
  83           0           0           0           1           4
  84           0           0           0           1           3
  85           0           0           0           1           3
  86           0           0           0           1           3
  87           0           0           0           0           3
  88           0           0           0           0           3
  89           0           0           0           0           3
  90           0           0           0           0           3
  91           0           0           0           0           3
  92           0           0           0           0           2
  93           0           0           0           0           2
  94           0           0           0           0           2
  95           0           0           0           0           2
  96           0           0           0           0           2
  97           0           0           0           0           2
  98           0           0           0           0           2
  99           0           0           0           0           2
 100           0           0           0           0           2
13

I can be on board for ā€˜linearā€™ or ā€˜not linearā€™ as the only two choices (assuming ā€˜linearā€™ has an impact - canā€™t wait to find out!)

Debate among different ā€˜not linearā€™ choices I find as amusing as WPPR v7.0 discussion :slight_smile:

2 Likes
14

If you are going to run a papa circuit event in future and have qualifying on 9 out of 12 pinball machines what type of point system in qualifying would you prefer?

15

For HERB, I think itā€™s fine for top scores to be worth more than lower scores (duh), but I just want it applied fairly across all scores and not just the first two scores!, lol, and to cherry pick a single example to support my claim ;-), In Atlanta, Jason scored 1,725,243,360 on GoT and Trent scored just below that with 1,714,151,980. Both, excellent games, but Jason was awarded 5 more points than the next player below him (Trent), yet Trent only received 1 more point than the next player below him (Chris). And that is asinine as hell!!

2 Likes
16

They all qualified . . . nothing to see here :slight_smile:

17

The problem with that is it is super dependent on the game scoring for the individual machine. Some machineā€™s scores will be flatter than others. Any kind of scaling based on raw score then creates an incentive to play machines that can be ā€œblown upā€ more easily to the detriment of flatter-scoring games. It also changes your play strategy from ā€œget a good scoreā€ to ā€œblow up or bust.ā€ Thereā€™s some of that now with HERB, but it would become more extreme with scaled scoring. I do agree thereā€™s an issue with cases like the GOT one, and Iā€™ve argued about it in the past. Linear is the one system that is least mathematically biased since it doesnā€™t care whether the big gap between scores, if any, is from #2 to #3 as above vs. from #5 to #6 or #1 to #2.

18

Added everything!

19

no scoring system is gonna help me get above that cut line in main!

20