Villain is site pro Ian Simpson - I know he's a pro as he has a custom avatar but I don't know a lot about his style other than he seems "reggish" at this final table, as one would expect.

Unibet Supernova 90+10 euros Blinds(4500/9000/1125) (Payouts 4000, 2700, 2000, 1460, 1020, 900 euros)
LJ: 360492.50
HJ: 272982.25
CO (Hero): 557832
BTN: 393285
SB: 231396
BB: 749012.25

Hero dealt: Q5

Preflop (pot 20250)
Hero raises to 22500, SB folds, BB calls 13500

Flop (pot 56250) - effective stack 534207 J45
BB checks, Hero bets 22500, BB Calls

Turn (pot 101250) - effective stack 511707 J45 - 2
BB checks, Hero checks

River (pot 101250) - effective stack 511707 J45 - 2 - 7
BB checks Hero ????

How much do we think he caps his range by checking here? Is it enough for us to make small value bet?

7 is not a good card for you as it completes straights as well as any hand he could be calling you with OTF 67, 87, 37 not to mention any other hand that maybe beat you. We have no blockers and it seems like a very thin spot to value bet. I'd probably check it back.

Pre i'm not very sure about.
As played i am probably vbetting small(1/3rd or so) and folding to a raise. He can also have weaker 4s/5s and might even hero with an Ace.
My 2 cents.

have to agree that pre isn't great unless the villains in the blinds are super nitty

^ I'd say Q5s is about the widest I would go so it's borderline for me too. Particularly the BTN and SB were quite tight passive, not 3-betting much (nor was Simpson in the BB excessively). This is what passes as the "Sunday Major" on this small poker site so there are lot of people who satellite in.

I've said this elsewhere to you, but Q5s is a clear check back on the river.
The board runout is so much better for villain's range (given that you checked back the turn), that your river value bets are all pretty thin apart from three combos of 77 that binked a 2 outer. You can't bet that river at a high frequency when villain has so many nut combos, and you've got almost none at all. (You never have a straight or a set when you check the turn. Villain can have all of those hands).
You can bet TT-88, K7s, 87 for thin value when you pick that small sizing. When you do bet the river for thin value, you're mostly targeting 7x, 66 and the better 5x. (i.e. almost no 5x/4x that Q5 beats should be calling. Indeed hands like 85s/53s/64s/43s can be turned into check-raise bluffs).

On the river, villain should be checking a bunch of very strong hands that want to check-raise. His river leading range is mostly two pairs and all the Jx, as these hands get called by your bluff-catchers. Better hands than top pair (86s, A3 and some of the sets) want to get an extra bet in the pot by check-raising. In other words, villain is far from capped when he checks the river*. He bets his medium strong hands (Jx) and some air, but checks nutted combos (straights, sets) to check-raise, and also checks weaker pairs that can win at SD if you check back, but that can be turned into xr bluffs if you bet.

* At least in theory. Most players are incapable of checking the nuts OOP on the river if the turn checks through, since they think they have to try for value, and this problem means that their checking ranges frequently are quite weak/capped. There's actually more EV in checking to induce a last minute bluff, or a thin value bet, since that allows a check-raise with a polarized range that prints money.

I agree that the whole thing depends a lot on whether or not villain caps himself with that check.

I'm quite new at this level having moved up from the 10E-25E level (still play there when the 50-100 SNGs aren't running) and with the player pool I'm used to, villains (including good ones) would not tend to check here with 7+ because the player pool as a whole doesn't bluff missed rivers enough. This hand has taught me I've got lots to learn now I'm swimming with the sharks in the big sea

Results and Villain comments during and after hand (he was streaming it)

Yes it does tend to cap his range. Most players will not check there hoping you bet with a value hand. He most likely does not have a strong hand himself. So you can eliminate top pair good kickers +. But that still leaves him with weak Jx combos, 7X, TT, 99 all check as well.

But him having a capped range does not mean you can value bet. The other 1/2 of making a value bet is asking yourself, what hands will he call my bet with that you will win at showdown. You have to win 50%+1 of the hands he will call with for a value bet. How many 4x hands or 33 call here?

Preflop though, I'm ok with this steal attempt. Q5s is in my steal range.

Looks like a cool spot for some overbets.

Well in the commentary villain says he doesn't always call with a 4 but it strikes me then that he must be exploitably overfolding at some point during the hand.

Obviously he is dealt ATC and all he has done up to the river is call a 2.5x open from the CO in the BB (with antes) getting 4.17 to 1, and to call a 40% c-bet ("min defence frequency" 78%) and now he is facing a 31% bet on the river (min defence frequency 83%).

I'm aware of the theoretical problems with min defence frequencies but the above is a good illustration. Perhaps he is defending a lot more by bluff raising so his range is stronger than I estimated but at least in game I was expecting to get hero-called by stuff like A high.

I think I might need to get a solver ......

This is exactly why I defended the steal with Q5s. With position and initiative, the blinds have a hell of a time making this unprofitable.

But anyway, from the BB perspective, he is supposed to call with MDF when your range is balanced on the river. If he has reads you are unbalanced, he should not be calling with MDF, but should adjust that number based on your range.

Now flipping that, he needs to call 76% when your bluffing range is balanced, ie, your bluffing range is 76% as well.

However, your are not supposed to call with MDF if the reads on the player or the run out indicate he cannot have that many bluffs in the range.

If you didn't have reads on him, it is likely he had no reads on you. And against unknown amateurs, there is one thing that is true about most of them. Almost no one bluffs the river as often as they should.

Knowing that, you should never use MDF on the river against an unknown amateur.

I take your point, but I've just seen I messed up calculating his min def frequency.

He can call river if I have 81% or fewer value combos as his odds are 31500/164250).
His "min defence frequency" on the river would be 101250/132750 = 76%. Of course he doesn't have to call 76% of his range if he thinks less than 19% of my range is bluffs but that means he is exploitably folding too much.

3 minutes later when it went out on twitch, the audience thought I had air. Personally, getting better than 4 to 1 I'm snapping people off with A high in his position, but then I'm probably a suspicious calling station.

No, your bluff range has to be higher than 19%. The formula for bluffing frequency is:

Is f = s/(s +p) = 23.7%

Now if you look at the number of hands that you can value bet on the river, if you have say a total of 20 hands, you need to bluff another 6 hands for a balanced range.

Where s is size and p is pot at the start of the street?

That formula is correct to get the number of bluffs as a proportion of the value hands rather than all hands bet, so we would have slightly less than 5 bluffs if we have 20 value hands.

The formula to get bluffs as a percentage of the total would be
s / (s+s+p) (or alternatively the same formula you state except where p is defined as including the amount after we have already bet) but it should be roughly 5 out of 25.

Using the first way, imagine traditional English odds (e.g. something to one, not including the stake). When we are indifferent to wagering on a horse at 3 to 1, we think it will lose 3 times as often as it wins so it will win 25% of the time.
We are offering him the chance to wager 31.5K in order to profit 132.75K, so the losses for him (our value hands) should be in 132.75/31.5 proportion to the wins for him (our bluffs)

The other way to think about it is with decimal odds, when we are indifferent to wagering on a horse at 4.00 it is because we think we win 1 time out of every 4 trials, so our wager needs to be 1/4 of the total amount that gets shipped to us if we win. In this case, to be indifferent villain needs us to have a bluff 31.5/164.250 of the time.

To test this, note that he has a slightly profitable call with 5 bluffs and a much more profitable call if we have 6 bluffs.
Running the call 25 times, 5 bluffs 20 value,
Villain puts 31.5Kx25 into the centre, is shipped back 164.250Kx5, total
821.5 - 787.5 = 33.75K profit or 1.35K per hand.

Running the call 26 times, 65 bluffs 20 value,
Villain puts 31.5Kx26 into the centre, is shipped back 164.250Kx6, total
985.5 - 819 = 166.5K profit or 6.4K per hand.

Either way - he doesn't need us to have many bluffs to have a profitable call with his bluff catchers, which is why I thought he would call really wide.

It's too thin, your villains need to be pretty horrible for this to be good. Literally nobody bluffs with your line so nobody will hero call you here.

no, it really doesn't

I'd be interesting in hearing more of your reasoning. One reason I think this river is a good candidate for overbets is I think hero's range can credible include straights and 2p.

Was he really considering folding sixes pre????

1. why would you possibly want to bluff Q5 here?

2. How does hero's range contain staights and 2pairs? 86s is 4 combos and barrels turn 95% of the time with the double gutter. What 2pairs do u have? 57s? J7s? You can maybe scrape together 6 combos if you're really ambitious, but not really. So yeah overbetting is horrible.

^Yes I also don't see how Hero reps any straights. And yeah there are a couple backdoor hands J7s/75s/77 but problem with those is they're gonna block V's hands that could consider hero calling.

More broadly as for our overall strategy, if we want to play B/X/B line here with 88-TT, maybe some Jx, then we probably shouldn't split our range small/big only to have extremely capped small size otr. Especially where Villain likely has all manner of straights getting to river (A3o, 63s, 86s).

It's such a low frequency of having nutted type value hand to OB for Hero that I don't see enough EV potential of splitting range that would overcome the potential complications/exploitability of having two sizes.

It does look like that on the video, but I think he is just talking through his thought process for his stream viewers and is never actually folding 66 here.

I discovered from watching the video that he already considers the other two opponents to be weaker than me (this site has name changes and no HUDs so I assumed he would still be readless but as it turns out, he's single-tabling and he has better reads than he might otherwise), so he might just prefer to get his chips from them and cut me up later in the HU match (which didn't come as I was 4th). Still think he's not folding 66 though.