On second thoughts, the probablilities really shouldn't have been squared (too many damn brackets ;-)), so the real figure is var(x) = 1/5*(2buyin^2) + 2/5*(-buyin)^2 = 6/5*buyin^2 but that doesn't affect the fact that the std. dev. is a linear function of the table level.

heh, yesterday I didn't feel like boring people with the math

I think it's hard to come up with a theoretical equation since so much is skilled based. It's more about psychology when people think the risk reward is worse on higher tables.

big table = win big, lose big
lower table = win peanuts, lose peanuts

apart from dom and kills, payouts are zero sum so what the winner wins is what the losers lose

for a 5 player 500, it would be +500 +0 +0 -250 -250. multiply by 4 for 2k and by 10 for 5k. An average player would get each position in 20% of his games. so 1/5*500 + 2/5*0 +2/5*-250 = 0

The theoretical risk reward at any table = the same. If you're "good", your risk reward will be in your favour since you will get less than 40% of 4th and 5th and more than 20% of firsts. Actually the biggest gains to make are in to avoid 4/5 and get 2/3 instead since even the most skilled players get a win% of 19-23%. That has to do with the fact that they usually play 2k/5ks where the competition is more fierce so you won't get many easy wins. If you're not that good, your risk reward will be bad on any table anyways.

I would say that dom is pretty much zero sum too and fairly linked to the basic points for winning and losing (wins = usually good dom, losses = bad dom). a skilled player will always find a way to maximise dom and to minimise losses in case of a bad game.

what I said about kills before is still valid. kills/game is pretty much stable on each table limit, it's more linked to your playstyle (do you go for kills or not). Probably you might get slightly less kills on higher tables because people know how to limit point loss so they won't drain dom like a noob. On the other hand, some people on 2k and 5k are nicer with giving kills compared to 500.

anyway, roughly 2 or 3 kills happen in each game and kill bonusses are 100,200 and 500 /kill so you can easily figure out where there is the most reward in getting kills

In conclusion: it's all about psychology. The risk reward is better on 5k or 2k, assuming you have the skills to play well. If you're average skilled and want to improve, it's probably best to stick to 500s to earn some safe points first and use the 2k to not care about point losses and try and learn something there. 1k tables won't help you in any way, you are simply going to face the same problems at 2k later if you don't know how to improve your game :)

"It's more about psychology when people think the risk reward is worse on higher tables."

Theoretically, risk is not the expected value of a game (which, as you correctly pointed out, is 0), instead it's linked to the deviation around the mean value, and is usually symmetrically defined, like e.g. a game which gives you $100 if you toss a coin and the result is heads but takes $100 from you if it's tails has a mean value of 0, but risk greater than 0 (var(x) = E(x^2 - 0) = 1/2*(-100)^2 + 1/2*100^2 = 10000; sqrt(10000)=$100). Risk is greater on the higher tables simply because the stakes are higher.

Even if the definition of risk is not symmetrical, like the concept of "Value at Risk" in portfolio management, it's linked to the risk of loss (say the maximum loss at 95% probability), and depends on the stakes.

Risk surely is greater on the higher tables but I'd say the reward goes up with the same factor.

Someone should note down 100 results from 5 player 500 tables and do the same for 100 2k or 5k tables. I'm pretty sure that the average payout for 1st,2nd, 3rd, 4th, 5th would be very similar, except for a factor of 4 or 10.

The variation comes from luck (or randomness) and skill (better players will have a better ppg on the higher tables) whereas less skilled players would drop in ppg going from 500 to 2k)

In fact, if you look at top 100, most people have a positive ppg (points per game). I would assume that most of these players mostly play 2k/5k compared to 500. Maybe not all but not that important. The point is that most have a positive ppg, even when ppg currently does not add the kill bonus.

So basically these people beat the expectation where you theoretically should get 1/5th of 1st, 1/5 of 2nd etc on 5 player games + 1/6 of 1st on 6 player and 1/7 of 1st on 7 player etc

tl dr again, still haven't heard anything that would make me change my mind about leaving the situation as it is (no 1k tables) and encouraging players to learn how dom works on 500s and occasionally try out 2ks when your point total reaches 5k.

If you want a safe way to get a white medal, you can easily do that on 500s. Not impossible for blue either but best way for that is to play 2k's or higher. Makes perfect sense to me that there is a "skill barrier" like that.

wow, thats a lot of over my head math :)

Essentially yes, my issue is that the tables don't fill reliably, which leaves payout only for first..

Maybe adjusting the payout levels for less populated tables would have a better overall effect than adding 1k tables, but I figured adding 1k tables would be less work / wouldn't impact how other tables played, and would "solve" the problem by being populated fully more regularly.

didnt see jurgens last reply before responding somehow.. definitely not trying to flog a dead horse, was just a suggestion from my perspective :) cheers!

oh no worries dnaa, I partly understand what your problem is

maybe a small change to the payout structure could help but the idea behind "winner takes all or most" was to encourage people to play for first more.

maybe 10-20% of the winnings should go to 2nd on a 5 player table too

the best solution would be more players on 500 and 2k though :)

Jurg, point of my post was about players playing in the 'european morning' dont want to play 2k. Simply its imposible to gather 5 players on 2k and 5k dont want even comment, so basicly when i come in the morning i cant play main. We have 3 solutions: 1. remove point limit from 500; 2. put limit for 500 on 15 or 20k and after that force them to play 2k; 3. make 1k tables without limit. Its fucked up when you log in and cant play... :(

given Fatman's 3 options, I'd prefer 1 or 3 (clearly).. option 2, might be OK if there was some adjustment to point dispersion. As it is, I play tourneys and 500 tables to build 15-20k over a few days, then play 2/5k for a day, and am back at 10k, fairly consistently. Being forced to play at higher level games may increase the rate of this "trickle up economics", and potentially frustrate / lose players. If the proposed limit were more in the 50k area, I think it would be a more acceptable option.

dnaa, all people who have 50k play 2k or 5k, that is point, if you move limit to 50k you didnt do nothing. 50k is top 20, top 20 play those tables anyway, problem are those guys who stop to play as soon as they reach 20k and little dot, i was like that. When i started old point system was aplying so then you needed like 12,5k for like 90th. I would play with main until i dont reach 14k so i get secured that i will get trophy and then i started to play this account. As you see this account also have white dots and that means if i played main instead of being afraid and playing alt i would acumulate blues from the start. A lot players have that problem in the head cause they think we are some 'raiding party' who just wait to kill them, after i played few months there to prove i wont stab you start to be treated diferently. You have ton of players who plays 100 last day of the month and they have like 13k, they need like one good win on 2k or on 5k for trophy but they forcing to play hundred of 100 table games to get dot. those kind of players we need to force to grow up...