PokerTracker

I've created a stat for flop donk bets, haven't been able to really test it yet but it should be fairly accurate.
I created two columns the first one is a count of donk bets and the second is opportunity to donk bet.

cnt_f_donk:
sum(if[holdem_hand_player_statistics.flg_p_face_raise AND
(holdem_hand_player_statistics.flg_f_cbet_opp = FALSE) AND
holdem_hand_player_statistics.flg_f_bet AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position), 1, 0])

cnt_f_donk_opp:
sum(if[holdem_hand_player_statistics.flg_p_face_raise AND
(holdem_hand_player_statistics.flg_f_cbet_opp = FALSE) AND
holdem_hand_player_statistics.flg_f_saw AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position), 1, 0])

Donk Flop = (cnt_f_donk / cnt_f_donk_opp) * 100

There seems to be a problem when the SB raises the BB but doesn't cbet , the BB is incorrectly getting a donk bet in this situation if he bets. I'll try and fix it tomorrow and post.

There are some donk-bet flop stats by Kraada in the Repository.

Yeah I saw those but I think they are a little off. His stat would filter out the situation where the donk better 3 bet preflop and just called a 4bet. By using flg_cbet_opp these situations would be included as a donk bet. I think I have the problem with the SB vs BB worked out but haven't been able to test it yet. I believe if I add pf agg. pos !=8 this should take care of it since you can't donk into the SB.

These have been updated, we found that problem and fixed it - I think the original (incorrect) one has now been removed from the repository - anyway, you need to redownload the latest dated one.

It looks like the same ones are in the repository, these will exclude the times someone 3bets and just calls a 4bet and donks into the raiser(or 4bets and calls a 5bet etc.). They also look like they will have the same problem I was having in blind vs blind situations. I've changed mine to where I believe it should be correct in BvB situations as well now.

cnt_f_donk:
sum(if[holdem_hand_player_statistics.flg_p_face_raise AND
(holdem_hand_player_statistics.flg_f_cbet_opp = FALSE) AND
holdem_hand_player_statistics.flg_f_bet AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos !=8) AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position OR
holdem_hand_player_statistics.position = 8), 1, 0])

cnt_f_donk_opp:
sum(if[holdem_hand_player_statistics.flg_p_face_raise AND
(holdem_hand_player_statistics.flg_f_cbet_opp = FALSE) AND
holdem_hand_player_statistics.flg_f_saw AND
holdem_hand_player_detail.val_p_raise_aggressor_pos != 8 AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position OR
holdem_hand_player_statistics.position = 8), 1, 0])

donk % = (cnt_f_donk / cnt_f_donk_opp) *100

cnt_f_donk_opp was wrong it should be:

sum(if[holdem_hand_player_statistics.flg_p_face_raise AND
(holdem_hand_player_statistics.flg_f_cbet_opp = FALSE) AND
holdem_hand_player_detail.flg_f_open_opp AND
holdem_hand_player_detail.val_p_raise_aggressor_pos != 8 AND
(holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position OR
holdem_hand_player_statistics.position = 8), 1, 0])

pkrlvr wrote:It looks like the same ones are in the repository, these will exclude the times someone 3bets and just calls a 4bet and donks into the raiser(or 4bets and calls a 5bet etc.). They also look like they will have the same problem I was having in blind vs blind situations. I've changed mine to where I believe it should be correct in BvB situations as well now.

Yes, you're right - it does miss the 3-bet / call a 4-bet case.

However, I don't think yours is quite right either.
This bit:
... AND (holdem_hand_player_detail.val_p_raise_aggressor_pos !=8) AND ...
means that it won't count a donk if the PF raise aggressor is in the BB.

(I haven't tested it yet, though..)

I'll have a further look at this again later on.

Right - I think I've got a better version..

cnt_f_donk =
sum(if[ (holdem_hand_player_statistics.flg_p_face_raise OR (holdem_hand_player_statistics.cnt_p_raise>0) ) AND holdem_hand_player_detail.flg_f_open AND NOT holdem_hand_player_statistics.flg_f_cbet_opp AND (holdem_hand_player_detail.val_p_raise_aggressor_pos < holdem_hand_player_statistics.position) , 1 , 0])

There was a raise preflop - either by you or someone else (you made or faced a raise).
You opened the betting on the flop, but it wasn't a cbet opportunity, so you weren't the final PF aggressor.
The PF raise aggressor was nearer to the button than you.

donk_opp is the same, except "holdem_hand_player_detail.flg_f_open" is replaced with f_open_opp.

Now, the problem about the blinds still exists in this - but this is because there is a bug in the "holdem_hand_player_detail.val_p_raise_aggressor_pos" stat - it sets blinds' positions as seats-off-the-button instead of setting them to 8 and 9 as it should. This will be fixed, so really the 'correct' version of this stat should assume this, but for now, we will need to include a workaround.

I'll come back to that in a while - I have to go out again now.

This bit:
... AND (holdem_hand_player_detail.val_p_raise_aggressor_pos !=8) AND ...
means that it won't count a donk if the PF raise aggressor is in the BB.

if SB=8 and BB=9 it should work. That bit there just tells it that if the SB was the pf aggressor a donk bet isn't possible by anyone. I added that because I was having a problem when the SB was the pf agg. but checked to the BB, if the BB bet in that situation it would incorrectly count as a donk bet.

If the holdem_hand_player_detail.val_p_raise_aggressor_pos stat isn't giving a value of 8 or 9 for the blinds it will be kind of tricky to work around. So at a full 6 max table it is setting the SB=4 & BB = 5 but if there were only 5 people at the table SB=3 & BB=4 ?