Buzzard's Article on Daily Fantasy Games (1 Viewer)

[(HULK)]

I won't post the whole article (which I found to be mostly insightful), but I will post a part that I am taking issue with. If its an issue (cause its paid content), someone tip me off and I'll remove the quote.

The third option I haven’t even mentioned are 50/50’s. At first glance these might seem like the lowest variance group because all you have to do is beat 50% of the players. It would seem that if we are some of the best players we should consistently beat half the people. However, after playing a few weeks you will see that this is easier said than done. Even the very best players will beat half the people only about 60-65% of the time.

What happens when we only beat 20%, 30%, 40%, or even 49% of the players? We lose all of our money! Compare this to a H2H game. In the H2H game we get exactly what we deserve. If we played 1,000 games and finished with a twenty percentile score we would win 20% of the time. If we finish in the eighty percentile we would win 80% of the time.

As you can see the 50/50 is great on weeks where you do above average but awful on weeks when you do below average. Since most of the games for even the best players are close to the 50 percentile range you are introducing a lot of variance by using 50/50’s, albeit with extra upside. But if we only played 50/50’s with large chunks of our bankroll we could go broke very quickly due to variance alone.

I just find this to be completely wrong. If you play 1000 games head to head and have a 20 percentile score, you'll still win 20% of the time. And you'd probably only win like 3 to 5% of the 50/50 matchups with this score. But what is missing here is that if you score in the 80th percentile, you only win 80% of your head to heads, but you will win likely 95 to 98% of 50/50s.

I played hundreds of dollars every week last year, and 50/50s were pretty much my bread and butter. If you routinely score over the 50th percentile, they're great games.

Which is better for you (50/50 or H2H) really comes down to what kind of points your team is going to put up. If you're confident your lineup is going to break 130 points (@ Fanduel), you should spam that into every 50/50 you can because you'll win almost all of them.

Anyways, I just completely disagree with this part of his article and wanted to point out this part because I'm sure I'm not alone in feeling this way. I am very happy to see more articles on the topic though, and eager to see the upcoming bankroll management article.

ETA: I may be biased because I once lost a H2H while scoring 184 points. Never would've happened in a 50/50.

 

[False Start]

Point is, these games are a waste for most people.

Sure some people win while others win big, but for most they just lose their cash. Its a tough science to crack because you are not playing against other peoples knowledge, you are playing against their bank roll where they buy up most the times to increase their odds.

 

[andrewd173]

Looking at the article, I believe the point was about risk of the game type. You can control that somewhat through bankroll, but it appears this article was just focusing on one aspect.

With the overly simplistic assumption that you put 100% every week into 50/50s, and you regularly score in the 60th percentile, it only takes one injury to knock you down to the 49th percentile and wipe you out. (Jake Locker last year killed me one week.) Whereas you would be down 1% (plus rake) with h2h. Granted the inverse is true, but again it appears to be an illustration of game mechanics.

Personally, I rolled with about 33% in 50s and h2h once I understood the risk.

 

[Ignoratio Elenchi]

What happens when we only beat 20%, 30%, 40%, or even 49% of the players? We lose all of our money!

And if you beat 50% of the players, you double your money (minus the rake, of course). I'm not sure why the author says this like it's a revelation - that's the whole point of the 50/50 format.

Besides, you don't lose "all" your money. You lose the money you entered in that particular contest, the same way you lose "all" your money when you enter a different format and don't win.

But if we only played 50/50’s with large chunks of our bankroll we could go broke very quickly due to variance alone.

I played pretty much nothing but 50/50s all year long last year, and I did great. You mitigate variance by entering multiple different lineups into a bunch of different contests - the exact same way you mitigate the enormous variance inherent in head to head matchups.

 

[Marauder]

What happens when we only beat 20%, 30%, 40%, or even 49% of the players? We lose all of our money!

And if you beat 50% of the players, you double your money (minus the rake, of course). I'm not sure why the author says this like it's a revelation - that's the whole point of the 50/50 format.

Besides, you don't lose "all" your money. You lose the money you entered in that particular contest, the same way you lose "all" your money when you enter a different format and don't win.

But if we only played 50/50’s with large chunks of our bankroll we could go broke very quickly due to variance alone.

I played pretty much nothing but 50/50s all year long last year, and I did great. You mitigate variance by entering multiple different lineups into a bunch of different contests - the exact same way you mitigate the enormous variance inherent in head to head matchups.

True, but it seems to me the variance inherent in 50/50's is clearly much higher than in HTH.

Let's say you enter the same lineup in all your contests and your team scores in the 40th percentile. If you entered all HTH contests you would expect to win on average 40% of your matchups just due to the varying skill of your opponents, correct? In 50/50's, I would think it would be pretty rare to win between 40-60% of your contests in any given week, assuming you are entering similar lineups in most of them. You are most likely going to win most of them or lose most of them.

If you are a good player, I think 50/50's are a great way to maximize your profits over the long term as long as you don't commit too high a percentage of your bankroll each week. You just have to be willing to withstand the variance.

I get your point about varying your lineups but if you do that too much you are probably hurting your long term profitability by submitting some lineups that have a lower EV than your more optimal lineups.