How to calculate ROI/expectation with BJ tournaments

Discussion in 'Blackjack Tournament Strategy' started by tucson1972, Jul 21, 2013.

  1. London Colin

    London Colin Top Member

    Understood.

    I had been trying to hammer home the point to tucson1972 that the only variables of any significance (in the absence of any skill-based calculations) are the amount of money on offer and the number of players competing for it. In that context you seemed to be pulling in the opposite direction by talking of rounds.:D
     
  2. London Colin

    London Colin Top Member

    All becomes clear...:D
     
  3. London Colin

    London Colin Top Member

    I can't speak about the norm in the U.S., but in the handful of land-based tournaments I've played in the U.K. that's been the case. Effectively a minimum prize pool is advertised to drum up some interest, and if more than this amount is taken in entry fees, then 100% of this is paid out. (I think that may even be a legal requirement here.)

    However, if the turnout is particulalry poor, and the casino managers do not mind generating some bad feeling, the minimum prize pool might prove not to be a minimum after all!


    And you will have seen variations on the theme online. While SitnGos and some MTTs will have a flat rake of 10% (or whatever), some MTTs have a guaranteed prize pool or have the pot sweetened with an additional amount, over and above what the players contribute.

    The most cynical example I came across was at GameAccount, where they combined a truly massive rake with an attention grabbing "£100 added" - https://www.blackjacktournaments.com/posts/42431 So what looked like a good deal was a vey bad deal on closer inspection.


    But that is knowledge after the fact. You can't know how many people will at be your table(s), whether you will get a wildcard, etc. in advance of deciding to play the tournament.

    In truth, you generally can't even know the number of players, so if the prize pool is capped, that is the most important thing to try and get a good estimate of.
     
    Last edited by a moderator: Oct 5, 2013
  4. S. Yama

    S. Yama Active Member

    $ cost and Ev

    I think tucson would like to see the exact EV of the GN blackjack tournament. To do so, other than the correct formulas provided by other posters in this thread, we should use concrete format with specific numbers of players. Also, we need to clarify if one attempts to just pay the original entry fees or rebuy and superrebuy, as well.
    From the link Colin provided it shows that there are 5 players per table in the semifinals.
    Let’s make a model, which may be not as exact as the one of the coming tournaments but should serve as an example.
    We plan on all rebuys, 4 out of 5 drawn names are present (in many tourneys it is much lower).
    The format could be as follow:

    1st Round
    162 players – 27 Tables (all tables have 6 players, two advance)
    54 players advance, 108 don’t advance
    Rebuy
    90-100 people rebuy, 18 tables, 6 or 5 per table (average 5.1), 36 advance
    Wild Card to 2nd round: 18
    total advancement to 2nd round 108

    2nd round
    108 players, 18 tables of 6, 36 advance
    Superrebuy 66-72 players, 12 tables, mostly 6 per table, 24 advance
    WC to 3rd round: 6 names
    total advancement to quarterfinals around 66

    Quarterfinals
    66 players, 11 tables, 22 advance
    WC to semis: 3 names
    total in semifinals 25 players

    Semifinal
    25 players, 5 tables of 5, one per table advances to the final
    plus 1 Wild card

    We can calculate the chances of advancing for the average skill player:

    to 2nd round:
    1st play 33.33%
    Rebuy (66.67x 39.2%) = 26.14%
    WC 40.5% x (18/58) = 12.6%
    Total chance 72%

    to quarters:
    1st play 72% x 33.33% = 24%
    superrebuy (72%x.66.66%) x 33.33 = 16%
    WC 60% x 6/82 = 4.4%
    Total chance 44.4%

    to semifinals:
    44.4% x 33.33% = 14.8%
    WC 85.2% x 3/112 = 2.3%
    Total chance 17.1%

    to final:
    17.1% x 20% = 3.4%
    WC 96.6% x 1/120 = 0.8%
    Total chance to make the final 4.2%

    As you can see making finals happens once in about 24 tournaments, which is slightly better than 6/162 (once in 27) because not all players rebuys and show for the wild cards.
    The total expected value is $696 ($612 from the final table prizes).
    The expected cost because of all possible rebuys is $1,073.
    So, for people who pay all the entries it is 35% disadvantage.
    However, I don’t know a single person who thinks that his skills are below the average player. For people who are better than average it is quite another story.

    S. Yama
     
    Last edited: Jul 23, 2013
    Ternamint and tucson1972 like this.
  5. London Colin

    London Colin Top Member

    Does it? All I can see is that one of the available wildcards is into the final. (Which admittedly I hadn't even noticed to begin with.)

    I can't see any specifics about the structure (no. of rounds, players per table, etc.). Maybe my IE browser isn't rendering the page properly:confused:

    (And it's also been suggested that there is in fact one less round than Tuscon believes.)
     
  6. tucson1972

    tucson1972 Member

    Wow! Mind blown Yama. This is exactly what I was looking for. But I don't understand how you came up with certain formulas. For advancing to round 2, of course I understand the .33 with initial entry but I don't understand these numbers.

    Rebuy (66.67x 39.2%) = 26.14%
    WC 40.5% x (18/58) = 12.6%

    With the rebuy where are you getting 66.67 and 39.2? I have no idea what these numbers represent. Same with WildCard 40.5%

    Are you playing this weekend at GN? If so I gladly buy you a drink/dinner and you can explain it then. Awesome.

    And Colin, I realize if you have the total number of entries that is the easiest calculation, don't have to nitpick with calculating the WC however I'm trying to figure out if a tournament is good before I enter. And I think breaking down the tournament like Yama just did does that to a T. I would be very comfortable playing in a tournament with these odds. But I appreciate your approach. It helped me see why my calculations were way off. You should visit Vegas and play the GN. It is lots of fun. Unless you re-buy twice. Then it sucks big time.
     
  7. London Colin

    London Colin Top Member

    That's what I'm saying is easiest done by dividing the prize pool by the estimated number of players (or entries, if accounting for rebuys - I'm beginning to confuse myself about which of those is applicable now! I think it depends on the circumstances.).

    The structure does not help you get at a better estimate. The structure is the same for everybody. Just treat it as a lottery. A number of people buy tickets, some of which prove to be winning tickets.

    Rebuys are a complication. I tend to think of each buyin I hand over as entry to a separate tournament. It's generally always worthwhile to rebuy, but in principle the assessment can be repeated at each opportunity to buy in.

    But as Yama mentioned, if you know you will be rebuying as often as allowed, hanging around for every wildcard drawing, etc., then you may be able to get more benefit from a particular structure than the average person.
     
  8. S. Yama

    S. Yama Active Member

    where do the numbers come from

    Thanks for the nice words tucson. I will not play GN this weekend but perhaps we may cross paths if you play some local (Vegas) weeklies.

    Where do the numbers come from?
    The average player advances to the second round 33.33% (2 of 6), then he rebuys only if he lost the first try – that’s 66.67%. He advances in rebuys 39.2% because we set the assumptions, as I wrote in the format description, that there will be on average 5.1 player per table (with exactly 5 it of course would be 40%). So his cumulative chances were 33.33 + 26.14 = 59.5%. If he advanced he doesn’t need wild card, but when he was unfortunate to not advance 100% - 59.5% = 40.5% then he is a part of wild card drawing. Since 90 people advanced from first play in the first round and rebuys there could be 72 people waiting for the drawing (162-90=72). We assumed that only 80% of them shows up (this may vary, I’d witnessed 15 names called to fill 5 spots, and sometimes all called are present), so 72 x 80% = 58. If they draw 18 names the chances for the wild card are 18/58th and “our player” will be eligible when he didn’t make to the second round through first play or rebuy – 40.5%. Thus WC 40.5% x (18/58) = 12.6%.
    And so it goes to the playing in the second round, he plays there 72% of the time, and advances 1/3 and not advancing 2/3rds. His superrebuy happens when he played in the second round and didn’t advance and his chances there are, once again, 1/3rd - thus we have:
    1st play 72% x 33.33% = 24%
    superrebuy (72%x.66.66%) x 33.33 = 16%
    60 players advanced from 1st play second round and superrebuy, wild card drawing have (162-60) x 0.8 = 82 players eligible, he didn’t make to the third round 60% of the time, so
    WC 60% x 6/82 = 4.4%
    ...and so on.

    The simple formula: total prize pool divided by the numbers of players works great for simple tournaments formats for average skill player.
    Even for better than average player it can be easily adjusted if he knows how much better he is for the whole tournament with that specific format (and has large enough record to justify it).
    However, if there are rebuys, and especially “superrebuys”, where only players who advanced to a higher round can use them, or there are “autoqulifiers” put in by the casino into higher rounds, or wild cards are not evenly distributed because they are earned by bigger gambling or for achieving specific goals, then analysis like mine above can present approximate value of the tournament. This should be adjusted by using specific “better than average per round”.

    Colin, the GN link has prize structure showing prizes for the finalist (1 to 6) and places 7 to 26 winning $500. That means there will be 5 semifinal tables of 5, one per table advances plus one wild card, and the rest (4 each on 5 semi tables = 20) gets semifinal bonus.

    S. Yama
     
    Last edited: Jul 23, 2013
  9. London Colin

    London Colin Top Member

    That does sound quite likely. But I can't see a description of the $500 as a semifinal bonus. Couldn't the semi be (for example) 5 tables of 6, with last place at the table getting nothing?
     
  10. tucson1972

    tucson1972 Member

    Hey Yama, you live in Vegas? We could still meet up if so. I'm in town this weekend and other than the GN tournament I have nothing else planned.

    Colin, from experience everybody at the SEMI table gets $500. The structure Yama put down is very accurate.
     
  11. London Colin

    London Colin Top Member

    So is it also 5 players per table, then, not the 6 you originally said?
     
  12. LeftNut

    LeftNut Top Member

    I have never seen more than 4 at a GN 100K semifinal table. YMMV.

    All semifinalists get the $500. I have also never seen less than 2 advance from any table/round except for the semi. They figure things out so that it works that way, they've been doing 'em a long time and know what's going on.
     
  13. tucson1972

    tucson1972 Member

    Colin,
    When I wrote this question I thought it was 6...having been to the SEMI's twice including last month I know for sure there were more than 4 people there. I would remember if it were four people. It may have been 5....if the numbers work out so that it was 5 I wouldn't disagree with that. There is another tournament this weekend and I'll do some detective work. I'll ask the staff how many players were entered, with and without rebuys and I'll make a point to see in each round the number of players playing for variance, etc. If there is anything you'd like me to ask/look into let me know. It is a great tournament, expensive but like LeftNut said...well done...and you get alot of rookies in the first round which is always a nice surprise.
     
  14. tucson1972

    tucson1972 Member

    So I guess the moral of the story is be careful of tournaments where the prize pool doesn't increase with the number of entries. Based on Yama's breakdown if the average player spends $1073 x 162 players= $173,826 and yet the prize pool is capped at $100,00. Doesn't seem like such a great tournament after all however it is much better than the terrible numbers I was coming up with.

    What is an acceptable keep for a casino in your eyes? Like Yama mentioned, it requires a lot of manpower, blocking off of the casino to host an event like this so I don't expect them to payout everything.
     
  15. London Colin

    London Colin Top Member

    It's this inevitable degree of uncertainty that makes me question the value of this whole approach. We have uncertainty about the actual structure that the casino will employ if the maximum number of players enter. We don't know what that maximum number of players is. And we don't know how the structure gets modified for other numbers of players. There are just too many variables.

    Alas, it's a bit too far for me to travel!:)
     
  16. Ternamint

    Ternamint Member

    Great ides, insights etc...
     
  17. Monkeysystem

    Monkeysystem Top Member Staff Member

    With analyses like this one it's easier to work with probabilities rather than odds. If of six advance your probability is 1/6 or 0.167. Your odds are 5:1, not 6:1.

    Because of wild cards, re-entries, etc. the probability of cashing is the number of finalists divided by the number of entries.

    As far as how the format of a tournament impacts its value for a player with above average skill, you have to consider the number and complexity of decisions that the format challenges the players with. Also, the impact that decisions have on an individual's outcome is a factor. How many hands do you have to play to get to the money? How many complex decisions do you have to make to get to the money? How strongly does the format reward dumb luck? This is more of a way to compare tournament formats rather than a way to quantify some sort of luck factor for a single tournament. And don't forget how soft the player pool is. A high buy-in open tournament populated mostly by crushers will have a bigger luck factor than one in which most of the players are clueless about strategy.
     
    LeftNut likes this.

Share This Page