The EV number of rounds for 2 player ladders was so awful I didn’t both putting it up.
The problem with ladder formats as it pertains to EV is that there’s a crazy number of byes. You could have 500 players but when you get down to 4 players you already know that 1 of those players only has to play 1 game, so his value accounts for 25% of the EV of the format.
I’m sure people smarter than I can discuss how the math works out beyond that point but as you move down the ladder the expectation for that player to win the event keeps decreasing.
You could take the middle road here and list approved formats as a suggestion for up-and-coming tournament directors, with a note that if they’re thinking about something new then they can email and ask if it would be kosher. Lower the bar for new tournaments but still allow innovation.
If you read the 2016 TGP guide we say exactly that
The TGP Guide lists “approved formats” for both qualifying and finals. Outside of that we’re definitely interested in hearing from organizers that have some potential new formats that could be added to this list.
Ok, I think I get it now. Its sort of a weighted average of how many games that final four players would have been expected to play. That makes sense when playing towards a final four group.
I suppose then a ladder that was played exactly like the bowling tournament example where matches would stay head to head 2 player games until the end would just always count as 2 rounds based on the top player being 50% probability of winning in 1 round, 25% in 2 rounds, etc. correct?
So I guess another question would be where does the line get drawn on the number of byes allowed before something starts being considered a ladder?
Right now I have an event where the top four qualifiers go into a small ladder/bracket of best of 3 matches for the finals, starting with 4v3, winner plays 2, winner of that plays 1 for the title. Its basically the same as a single elimination bracket with single bye and double bye for the top 2 qualifiers. If I did that in 2016 would it be considered a single elimination bracket with byes or would it be considered a ladder? (Sorry for pestering about this format)
We plan on adding common bye methods to the TGP Guide. For organizers that want to do something outside of that box you can just email us and we’ll work it out.
I know Stewart ran it for me. It based on Circuit Final with the final 4 playing “one game”, so then you can adjust from there for whatever the final 4 actually is.
I was SHOCKED at how the math worked with ladder. Was expecting the EV to grow much quicker compared to what it actually does… But you know. … Math
Yeah I plotted it out in a spreadsheet for the head to head ladder style and after about 10 rounds the values are so small its not even worth adding it up anymore.
For any tournaments that have multiple paths of qualifying for the
finals, we will take the SHORTEST of those paths when counting
meaningful games played for that portion of the tournament. This is to
try and stop the confusing formats where organizers offer a way for
players to qualify for an event easily, and then have some subgroup of
players also battling out for spots by using a longer process. Couple of
examples would be taking having 5 machines and allowing the high score
of each machine a spot in the finals, while also giving players a chance
to advance to the finals based on their play across all 5 machines.
This will count as ONE meaningful game played because of that potential
path for players to reach the finals rather than 5.
This is inconsistent with the Expected Value used elsewhere. It’s actually easier to calculate an Expected Value for the [average] number of games a Finalist will have played to qualify in those cases than doing the math for some other formats, e.g. strikes. If you use EV for most of the components, you should use it for all of them where the math’s not nasty, and it’s not nasty here.
You could do a format where people play matches and the top 10%-50% of match play records go to an indirect “compare scores” format, e.g. the qualifiers play 10 machines and get ranking points [10-9-8-etc. or 100-90-85-84 etc] based on their scores’ positions among all qualifiers. Might be worth a try …
One way to get more qualifying games in is to count each player’s best 2 or 3 scores on a machine in the qualifying bank towards qualifying. Am I correct in assuming that the “no more than 3 games per machine per round” rule also implies that counting multiple scores this way in qualifying would be capped at 3?
True. Maybe miniature golf is a better analogy. For example you have a hole that has a tricky shot blocked by a windmill, where if you miss you hit the windmill and end up in a bad position, or you can take the safe shot around it which won’t get you a hole in one.
if you’re way behind you need to take the risk, if way ahead take the safe shot.
“I’m only behind by a stroke here, do I hit a 3 wood over the water to try and make eagle on this 580 yard par 5 (and win the tournament), or do I safely lay up to 120 and wedge in close for an easier shot at birdie and a tie” is pretty much the same thing as your windmill analogy.
@pinwizj: Just to clarify, for a league or event that is entirely based upon direct play (whether 2, 3, or 4 players per direct-play match grouping), there is not a need to have a distinct finals/playoffs portion in order to be IFPA-endorsed and earn WPPRs.
As annoying as it is to me personally for leagues to submit ‘regular season’ results, and then submit ‘finals results’ separately . . . under v5.2 rules you can still do that if your regular season consists of direct play.
Yep . . . since we don’t give any tangible benefits between calling something a “league” or a “tournament”, any organizer is welcome to put together a competitive event that results of playing a bunch of games and crowning a winner.
Some leagues submit results every single meeting. Some submit end of regular season, and then submit a separate event for their end of season “tournament”. The only thing we technically care about are the games played counting towards TGP are only based on the results being submitted.
This is where we’re seeing leagues modify their season to be as short as possible while grading out to 100% TGP as quickly as possible. In our Chicago league we don’t particularly give a shit about WPPR’s, so our season consists of 30 games played in 4-player groups, followed by an 8-player best 2-out-of-3 double elim playoff.
For us that’s 75 games played towards TGP per season, so we’re literally wasting 2/3rds of the games played during our “season” if you’re talking about designing a league for WPPR efficiency.
