I did research some time ago on this to provide a fairly simple approach for a math-based answer to multi-tabling quantity. It basically formalizes the usual suggestion to try different table quantities and see which one does best. Here is a repeat of a previous posting.
I define a table effectiveness factor (TEF) as the fraction of win rate per hour you retain each time you increase the number of tables you play by one (0<TEF<=1.0). Then with a normalized win rate per hour of 1, the normalized win rate per hour for n tables is n*(TEF^(n-1)).
To estimate TEF get the win rate per hour for playing 1 table, say W1 obtained over several days perhaps. Do that for several other table quantities, so you have W1, Wn1, Wn2 etc.. Then calculate TEF for each n as
TEFn= [(Wn/W1)/n]^(1/(n-1))
Order the TEF’s and choose the middle one (the median) as your estimate.
Example: You have data for 4 table quantities (too small a sample but okay for this example). You know from tracking data that for one table your WR/hr = 10. In several sessions, you played two tables and WR2/hr = 18.5. For another few sessions, you played 4 tables with WR4/hr = 29. For 8 tables win rate/hr was WR8/hr = 30. The corresponding TEFs using the formula shown above are as follows: 92.5%, 89.8%, 86.9%. The median value (rounded) is 90%. This type calculation should be done periodically for it is likely that you will improve your table effectiveness as you gain experience or perhaps the playing environment has changed.
Now once you have an estimated TEF, use the following chart (with interpolation) to determine how many tables you should play. Instead you can use the formula to see which n maximizes your win rate.
Table Effectiveness Factor | Optimal Number of Tables |
0.95 | 19 |
0.90 | 9 |
0.85 | 6 |
0.80 | 4 |
0.75 | 3 |
0.50 | 1 |
The model is simple and subject to question but it’s a start. Also, another factor to consider is that if more tables make you much less effective, then you might be better off playing fewer tables at a higher level.