Tournament extension question.

I’ve never played this type of format so not sure what it normal. I did understand the 1st extension as it was obvious that we had more time to play additional games, but the second extension would clearly have players starting games after the time cutoff.

Edit: see you saying it was 16 games, then 20, then 22
This would change my estimation for last time that the game count could have changed was at or before 4:11. Still not a wildly ambitious change to game count trying to fill the time to stretch to 5:00.

My time zone is central, so I think Matchplay displayed the time differently (I see ending at 4:35, but guess it was 5:35). Still not something that is uncommon or somehow misconstrued. I’m not convinced there were multiple extensions. Going much further than this I’d be suggesting again to bring the TD into this conversation or have the discussion yourself. Can report back here if you’d like if choosing the latter.

The way I see this is MaxMP set up with some 18 to 20 games. Went fast enough by game average time that it was changed to 22 games to fill the time limit. Last games dragged on since the matchups were getting congested with few unplayed opponents. There for sure could not have been any “extension” at 5:00 as there were only 5 remaining matches. If there had been a 2nd late extension the latest it could have been was at 4:36 as there would need to be at least one game per player remaining. Extending from 21 to 22 in that case. Even then the average game time was less than 9 minutes, which as this is a new format for most everyone could honestly be read to think there was still plenty of time for another game per player.

I’m not saying that it was a perfect run for the format, but I am saying I doubt 100% that there was anything done to create an unfair advantage to anyone.

1 Like

Hi, hello, TD of this tournament here. This was not a flip frenzy. The new max matchplay format does does not have a time cutoff , only a match cutoff. This is how it preserves fairness via equal number of matches vs the Frenzy.

During tournament announcements, I specifically mentioned that we would try to end around 5pm but that it wouldn’t be an exact science because it’s based on matches, not time. I had a friend there who was counting on leaving around 5, and I was very careful to mention this (actually multiple times) so as not to give him the false hope that he might be able to leave at 5 on the dot. You can set the time for a frenzy, and I have never ever changed the time limit of a frenzy once established. But you cannot set the time on a max Matchplay.

The first limit was set at 15 because, like most people, I didn’t know how long 15 matches would take and wanted to be conservative. When I realized that would not be enough, I increased to 18, then 20, then 24 over the course of the day to try and time it out as precisely as I could. I actually realized 24 was too much at some point so I took it DOWN to 22, which thankfully the software allowed me to do. Thanks Andreas. I absolutely DID NOT add matches after 5, and you left the tournament just several minutes after 5 along with most participants.

The tournament continued after you left for about a half hour. I did not anticipate the remaining few matches taking so long but like someone said there were fewer people to match with , so as many times as I pressed the button to make matches, math prevailed and said no. I couldn’t lower the number of matches at that time, because 75% had completed 22 matches and left, and everyone had to have the same amount.

Trying to guess how many matches = x amount of time is tricky, and being that only a few people even remained after 5:10, I feel like I did a pretty good job, but I would have chosen 21 matches to time it perfectly if I could go back in time. Everyone played the same number of matches and ultimately, though I enjoy a Frenzy more, this was more fair, and the feedback I’ve received about that aspect is significant.

I’m not sure who you think I was trying to advantage or disadvantage. I didn’t do well in the tournament and had nothing to gain from doing anything nefarious. I simply was trying to run a tournament with a brand new format until approx 5pm. It’s a freakin thankless job sometimes.

8 Likes

Thanks for your input here. Glad to see a TD’s perspective. I have a few suggestions for MaxMP to offer. Try to review the games played by all players often. I have ran (and played) in three. It seems like a good practice to help the math by making sure the game count between players doesn’t exceed two. I did this by stopping the creation of new games once this was identified. If the game is really going to drag and it stopped everyone from playing, then I would use the configuration to limit the max number of matches to about half what it was previously. Once the lagging player caught up the brakes could be released.

In some ways MaxMP is a bit more hectic than Flip Frenzy, but I think it is a great improvement even with a little bit of a learning curve.

1 Like

It was definitely more hectic, I tried to allow the queue/“waiting room” to get to 8-10 people so matches would be more diverse, and that was successful. Creating matches at every opportunity leads to lots of overlap in games and player matchups. I swear I created matches and then played one crappy ball of radical and there were 10 people waiting already. It requires constant attention. I’m sure at some point the Matchplay software will automatically manage the queue with input prior to start. Like… “create new matches when the number waiting is x”.

1 Like

Max Match Play is a very new format, so all of the TDs and players who are trying it out right now are helping to explore what’s normal.

I’ve been involved in two so far myself and have definitely learned some lessons, similar to what @Zaxxoff described.

Thank you @Lam for trying out this variation where the number of matches is adjusted to try to meet a timeframe goal. Based on seeing that match burndown, it looks like more experimentation is needed to figure out how to pull that off while also avoiding the “long tail of slow matches” problem that lead to overshooting the goal of ending around 5pm.

Maybe only ever increasing the target number of matches by 1 at a time would be better than jumping from 20 to 24. But waiting until everyone has completed 20 (or maybe 19) matches before increasing it to 21 would also lead to more people stuck in the waiting list longer and fewer games total. There are definitely tradeoffs.

I think this is a great guideline for when to click the Create games button. Based on my experience (and what I see in this burn down and in @Zaxxoff’s comments), I think it would be better when combined with one more rule. Also check for any players who are currently playing that already 2 games behind the people with the most games completed. If there are, wait until those players are done playing their match.

@haugstrup has mentioned he’s waiting for more data and feedback before iterating on the Match Play Events implementation of the format, so hopefully some TDs can try out that rule and provide feedback on if it helps make their tournaments better or not (e.g. no long tail of final matches; but still not too much waiting).

So true!

1 Like

I’m a bit more laissez-faire when running this format. I leave the tournament manager accessible to players on my tablet so that games can be created by the players (pushing the create games button). Then if I need them to stop I just put a note of the corner of the table that says, “Stop Creating Games”

A bit of a less structured format than most are used to, but I believe in the math and when it runs without rematches I think it is fantastically balanced.

I’m still working on trying to minimize arena redraws. Had a problem with that even for Flip Frenzy. The venue I’m hosting at has only 20 games and is open to the public so I don’t use every single game, which will pretty much guarantee some redraws. In our case I think it is a decent trade off.

One thing to keep in mind for her tournament is that she had an uneven number of players. You have to increase (or decrease) by at least two games.

2 Likes

I had three people creating games (with the guideline of : don’t do it until there are 8-10 waiting, but everyone was allowed to enter their own scores, without approval. Laissez Faire Lifesaver. Would be tough with new people, but this worked with experienced players.

I forgot about that! YES. The new guy arrived 40 minutes after tourney start and I learned a few things with the big red “wrong number of matches” notice kept displaying lol

Hahaha. Never again. :slight_smile:

Agreed! I don’t think I would try it now either.

Flip Frenzy is good for knowing an accurate ending time: x hours after start time plus however long it takes to finish the games that were started just before the buzzer.

Same with Timed Match Play: last round starts by x time and the tournament/qualifying runs only as long as it takes for that round to finish.

Max Match Play doesn’t naturally have that property and it looks like it’d be hard to accomplish it by adjusting the number of matches.

It’s tough even to know how many matches to plan for Max Match Play using a fixed number of matches if end time is an important factor. You ran 297 games across 27 players in 4.5 hours. So maybe it feels safe to announce the next tournament as “20 games of Max Match Play starting at 1 pm and finishing around 5 pm”. You expect 27 players to play 270 games in 4 hours this time. That feels safe based on this. But if 37 players show up you’re going to way overshoot your estimated time frame. Maybe once you establish a baseline it’s best to use this language for the IFPA calendar description: “number of games will be determined by player count at the start of the tournament” and then do some math like:

(2 players per game * 270 games in about 4 hours) / 37 players = maximum duration
(2 * 270) / 37 = 14.5
round 14.5 down to the next even number of 14

So you go with 14 matches per player that day. I’m not sure if the tournament timeframe would scale that way, but it seems like it might. 20 rounds with 27 players and 14 rounds with 37 players are both around the same WPPRs value too (as expected if they take the same amount of time).

The tournaments I was involved in were at two different venues and had these stats:

MPE Link Players Duration Time Total game count Long tail GPH
first 31 10 3 hours 155 None 51.7
second 20 12 3 hours 120 1 hour 40
yours 27 22 4 hours 30 minutes 297 40 minutes 66

I’m thinking that maybe games per hour (GPH) will be consistent at each venue, maybe with some adjustment for whether the long tail problem comes up or not. @haugstrup this kind of stat might be interesting to look at in aggregate after you have more data.

but then the issue of some players getting on long playing games vs fast ones can lead to gaps in # games played.
And also per IFPA rules you really can’t cut games just due them being to long.

I don’t think either of those concerns introduce problems with what I’m proposing.

What I will do next time is pick a set number of matches (probably 24) and ending time will be open-ended. I’ll still have a general idea of when it will end, of course. Just like I did this last time.

I’m lucky to have unlimited access to the venue for tournaments with no time constraints, so I have the luxury of not having to abide by those parameters moving forward, but I think everything y’all have discussed about finding the sweet spot is really important for TDs with limited or restricted access to the venue and greater time constraints.

3 Likes

Lovin’ the data comparisons here too.

@Lam

“I’m not sure who you think I was trying to advantage or disadvantage.”

I don’t think you were, nor did I imply that anywhere.

My frustration, and i shared this with you at the tournament, was that the games to be played changes that happened after the initial change were never communicated clearly to the players. Some players knew about it, but I only learned the changes had been made when I completed my 20th game and asked if I could leave.

I simply wanted to understand how this is usually handled as I found it really frustrating.
I now understand that these changes are not usual and I’ll make sure to have more clarity on what cutoffs are being used before atrendint an event like this.

1 Like

Yes, it feels like that’s the sweet-spot for creating new matches. What’s been helpful for me is that I’ve also deputized a few other people to keep on eye on the computer and said “When X people are in the queue then create matches”, so even if I’m stuck on a game it stills flows.

We also let people enter results from their phone which keeps the computer area clear.

We have a similar problem…I have found if you let the waiting list get long enough it minimizes Arena redraws.

That’s what we’ve switched to instead of trying to time-box.

We were able to record results from our phones and the queue never seemed to get too long, longest wait was probably 4 minutes.

@Lam
“I absolutely DID NOT add matches after 5, and you left the tournament just several minutes after 5 along with most participants”

I never said you added games after 5, I have no idea when the changes were made after the initial change to 20, but the games that were added pushed the start of games past 5pm. I think I left around 5:20, might have been a little later.

That was the point of my post, protocol for changing the number of matches played in a tournament. As a player I was never informed about the changes in game number, to 24 and then to 22, and why I was frustrated.