I did something very much like this for a few years. Ultimately though it's only as good as your inputs and your inputs aren't very good (as noted in the article).
On top of that the cumulative probability gets really bad because the predicted spread is never very high. The best you'll see is around 60-70%. .7 ^ 10 = .028, so while your expected value may well be positive if everyone else is using a poor strategy, your overall probability of winning is still low. At least in leagues with a lot of players (the one I was in had over 10,000).
That was definitely the end result. For my inputs the expected survival time was about 5.5 games. For a pool with 100 players, taking the conservative assumption that every match is a coin flip, you would still expect a few people to outlast you. Even if the payout was high enough that your bet was +EV, your variance is huge, so this was definitely more for fun than for profit :)
On top of that the cumulative probability gets really bad because the predicted spread is never very high. The best you'll see is around 60-70%. .7 ^ 10 = .028, so while your expected value may well be positive if everyone else is using a poor strategy, your overall probability of winning is still low. At least in leagues with a lot of players (the one I was in had over 10,000).