Re: MTT bankroll
I am just getting started as well, so I'll try to lend a hand from a beginner's perspective. I have similar questions, maybe we can get some experts to weigh in if we make it interesting enough.
My thoughts are that a BR of 40-50 times the buyin you plan to play most often should be sufficient. I am going with 2500 and I usually play 50s or 20/30 rebuys with a 100 once a week. I figure that a run of 50 times out of the money is extremely unlikely, and that ITM's can keep you afloat as you shoot for the real goal - final tables, top threes, and wins.
In general, I think the meaningful categories you can put in are:
Out first hour/Out second hour/etc..../Out on Bubble/ITM/Final Table/Top Threes/Wins. Maybe even track when you bust by level, that sort of thing might help you identify the phases of the tournament where you make the most mistakes or take the most risks (for good or bad).
Obviously, a final table in a 100 person tourney is obviously not the same as making one in a 600 person. So I plan on keeping records in several groupings based on tournament size.
I am trying to come up with an overall statistic to measure the rate at which I finish ITM or make FT's, etc. that is weighted for tournament size. In other words, if I play hundred tourneys with two hundred people where exactly twenty make the money, the average player makes the money ten times, the final table 5, and the top 3 1.5 times. I think a weighted ratio for each of those categories (where 1.00 = average, 2.00 = made twice as many as average, etc.) would be valuable. Maybe with a multiplier for buyin level, to reflect that making the money in a $300 means more than in a $10. If you are interested, I'll post the system i come up with once I have it worked. I think setting and hitting goals like "cashing twice as often as average" can keep you feeling good when you hit a run of not making final tables.
Of course, dollars in pocket is the stat that outweighs all others, but if you are the type to want a more thorough evaluation of performance according to key benchmarks, I think something along these lines could work.
Thoughts?
|