https://www.youtube.com/watch?v=aQ6wSLxt-BM

Check 03:00 in the video.

Pump faking could be useful to get a read of your opponent. What physical tells should you look for when doing it? In my experience, any sudden movements indicates weakness while when villain is sitting completely still it indicates a strong hand. What do you think?

What kind of sudden movements? The only one that I think could happen naturally is looking away (not up, which might be impatience). And that is more likely to mean weakness than strength, imo.

I think...for me.... when pump faked, I have moved involuntarily (when I was strong) to grab my calling checks.

Interesting hand. I might respond later about the pump fake. But prob more important than the pump fake here is just how weird his line is in this hand. Add the timing of the bet on the river (pretty fast, if the editing is at all accurate, which I'm told they try to keep pretty accurate for these kinds of spots) and it's very suspicious. (edit; just watched this again and it's not that fast, and it's entirely possible they edited a little bit out, which would make my point mute, and also of course possible that player bets pretty fast all the time no matter what, which would also make him hard to read)

If Ha had a flush on turn (basically what he's representing with the turn 3-bet) then the paired board on the river would presumably make him pause a bit to consider the situation and how things might have changed. If he 3-bet turn with a set and hit a full-house, it's often pretty common for someone with a huge hand to think a bit about the best way to get value and to convey a little uncertainty, too. This is why these fairly quick big bets on the river are weighted a bit more towards bluffs in my opinion.

Reminded me of this hand from 2007 WSOP. Similar bet-speed stuff and board texture:

I believe you're being too presumptive about bet sizing and bet speed. Villain and I would bet the same if I had Ah and putting Tran on Kh to Jh. Similarly, if villain flopped a set or two pair and filled up on the river, nothing would be different in bet sizing and pace. Bet speed can be attributed to a good player having a game plan going to the flop and executing it like an assassin on future streets. Remember back in 2007, it was proper to be hyper-agro with draws and applying pressure when in position.

After the initial bet on the flop, the villain is simply continuing his story by barreling off on the future streets. The rail and on-lookers are missing the meta-game and hand history between the two. Maybe the villian tried a similar bluff before. What IS LIKELY, is that both players lost their minds when the camera crew began filiming their table trying to show off how big their balls are and Tran simply held the better hand after the river.

Tran's call on the river is only a lucky guess since his hand is at best a bluff catcher. If Tran really thought he had the best hand and the villian was bluffing, why not raise the river bet forcing villian to fold and keeping your hand concealed. By showing down the river with such a light holding, Tran is saying to the rest of the table that he's willing to call large river bets and risk half of his stack extremely light. In turn, I lick my chops and imagine how good his chips will look added to my stack.

Analyzing bet sizing and bet speed by focusing on a few single case results is not a profitable angle.

This statement is backwards. Tran's call on the river is incredibly awesome because his opponent is so polarized that his hand is a fine bluff catcher.

I agree with you regarding how it could be a player with a game plan. I wouldn't be likely to rely on this read against someone who I thought was a strong player. Whenever I talk about reliable reads I'm assuming it's someone who's not very good. So in that sense I think we are on the same page.

And I also agree it is unknown what history is there, or even if the editing is accurate (although I've been told by a WSOP episode producer that they do try to keep the relative length of decisions somewhat accurate and wouldn't display something as a quick action unless it actually was a quick action). When I try to analyze televised tells where there is no history present, I try to just do it from a perspective of: what is this likely to mean in a vacuum, while keeping in mind in reality things are much more complex. It's not like I'm saying: this always means this. But if we can least say "this means x more than it means y' then that's useful at least in borderline spots.

Good reminder what a bag o Doosh KT is.