Tyson Streib Continuation Bet Article (May)

[Sherman]

[ QUOTE ]
Tyson,

great article, articulated well.

can you tell us: where you got the data set from? just curious. thx!

[/ QUOTE ]

[jogsxyz]

[ QUOTE ]
[ QUOTE ]
can you tell us: what was the size of the data set?

[/ QUOTE ]

The $10 set contained 662,148 hands, and from that there were 40,646 times where someone made a bet on the flop.

Tysen

[/ QUOTE ]

That seems low. Did you mean 40,646 c-bets?
How many times did the sample see a flop after a raise preflop?
What's the histogram for 2, 3, 4, etc way pots?

The defenders miss the flop about 65% of the time.
Here hitting the flop is defined as making a pair or better. Don't know how often one hits a OESD or 4flush. In two-way pots aren't defenders folding over 50% of the time? Do they fold more readily to a EP raise than a OTB or CO raise? Nobody believes the blind stealers.
The unexplored territory is what to do when your c-bet is called. When or if one should make a c-bet on the turn?

[MATT111]

Anybody has an idea of how the information provided may or may not be valid for cash games.
I tend to think we would have to bet bigger in cash games on average as stack sizes are typically larger than in tourneys.

[UnderThe Gun]

Am I missing something?

As I understood your article, it draws inferences about continuation bets by analyzing data drawn from all bets on the flop whether the bets are continuation bets or not.

As I'm sure we all understand, continuation bets are a specific subset of flop bets with characteristics that at least arguably make them more likely to succeed in taking the pot that most other flop bets.

So what's the basis for assuming the flop bet data says anything meaningful about continuation bets?

[Peter Harris]

Congratulations on the article. However, I wonder whether it's fallacious thinking when you do mathematical analyses of the success of c-bets on 0-gap flops and then equate 654 to JT9. It is more than likely that a c-bet on a JT9 flop where you miss will be less than successful than on a 654 flop given the nature of hands a generic player will call a preflop raise with.

There were no caveats to the inherent errors that will creep in to such an analysis when you bracket off vastly different flop situations into relatively simplistic analytical categories.

Whilst I concede that the article is useful and well written, I think a greater disclaimer that your simplistic bracketing ranges risk drawing too concrete conclusions would have been useful.

Regards,
Pete Harris

[Xhad]

[ QUOTE ]
As I'm sure we all understand, continuation bets are a specific subset of flop bets with characteristics that at least arguably make them more likely to succeed in taking the pot that most other flop bets.

So what's the basis for assuming the flop bet data says anything meaningful about continuation bets?

[/ QUOTE ]

Continuation bets' success are measured by whether everyone folds. Therefore if everyone folds n% of the time a bet is made, then that should tell you the success rate of a continuation bet.

[trojanrabbit]

[ QUOTE ]
Congratulations on the article. However, I wonder whether it's fallacious thinking when you do mathematical analyses of the success of c-bets on 0-gap flops and then equate 654 to JT9. It is more than likely that a c-bet on a JT9 flop where you miss will be less than successful than on a 654 flop given the nature of hands a generic player will call a preflop raise with.

[/ QUOTE ]

Actually, not really. Grouping helps keep the sampling error down, but I can slice and dice it into individual flops if you like:

All the 0-gap flops from JT9 down to 432 are essentially the same. If anything, the higher flops are more likely to succeed.

Tysen

[trojanrabbit]

[ QUOTE ]
What's the histogram for 2, 3, 4, etc way pots?


[/ QUOTE ]
Players seeing the flop when someone bet on the flop
2 = 24527
3 = 9595
4 = 4206
5 = 1695
6 = 487
+ = 136

Your other questions will have to wait 'till I have more time... [img]/images/graemlins/tongue.gif[/img]

Tysen

[jogsxyz]

The chart says against one opponent the c-bet works 52% of the for $10 limit players and only 33% for the bigger limit players. You conclude this means the c-bet is overused.

Players miss the flop 50-65% of the time. Maybe the bigger limit players are overcalling. In MOP chapter 19 the clairvoyant with only 2/13 chance of a winner is betting. Still the defender should only call pot/(pot+bet) of the time.
Even in the big limits the preflop raiser has a better distribution than the defender. The defender should be calling less than pot/(pot+bet) of the time, due to weaker distribution. The big limit players are playing macho and badly. The preflop raiser should punish him on the turn.

[threads13]

I have a question for you to speculate about Tysen.

In the beginning of the article you say that you think it is one of the most overused moves. So, what percentage of the time should one c-bet from looking at your data?