Boat Pente wedge analysis
Posted:
Mar 8, 2019, 7:37 AM

I'm going to collect here some of my (and other people's) thinking on various aspects of the wedge in Boat Pente. Included will be positions of interest and possible continuations for P1 to try to find a win from, branches of the wedge which I feel reasonably confident won't win as P1 and so on. I'm going to explore and notate the wedge beginning from the position of 1. K10 K9 2. N10 M9 3. L9 but I will also include translations of moves and lines of interest from 1. K10 L9 2. N10 N9 3. M9

Based on my discussions with Dmitri King, P1 can't diverge from the capture exchange sequence in regular Pente, otherwise P2 wins the game. However, this may not be the case in Boat Pente. Alternatively, P2 may need to find different lines in order to win in Boat Pente if P1 diverges. An illustration of this:

At move 16 if the above game had been Boat Pente, P1 could play play 16. M7 instead of M9 and then it appears to me that P2 loses the game on pairs instead of having the winning move of 16. ... L8. So, P2 would have to rethink that line in Boat to determine whether or not there was a different way to win it.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 9, 2019, 4:24 AM

Exhausting all potential divergences from P1's main wedge line is currently beyond the scope of my time to devote to Boat analysis, so we'll just assume for now that there may be other divergences worth exploring as P1 besides the one mentioned above: 5. M10. Even assuming that in Boat 5. M10 is a sure P1 win, it's unlikely but always at least within the realm of possibility that P2 has a different prior sure win move in the Boat wedge sequence. Perhaps 4. ... M10 or 3. ... L9 can win for P2 in Boat. It's unlikely, but can't be ruled out because:

1. We know that Boat Pente changes some Pente lines from P1 wins to P1 losses (and vice versa) even when the losing in Pente side has no captures. See the game in the post above above and also the game below for examples:

with 13. ... G10 or 13. ... H7 followed by 14. ... J7 winning for P2.

2. From K10 onward, either P1 or P2 has at least one sure win move with perfect play. Since P1 is restricted to playing K10 and then disallowed from playing certain moves (which moves are forbidden varies by game variant), restrictions can alter the balance of the Pente games in favor of a sure win for P2, if P1 is unable to play close enough to the center of the board on his second move and P2 has no restrictions on her moves. We'll assume for now that P1 has at least one winning second move to counter whatever P2 can come up with as a first Boat move, but while this assumption is highly probable it is by no means a certainty. Nonetheless, either 1. K10 + 2. x## wins for P1 against 1. ? y## or 1. ? y## wins against anything P1 can come up with. So, already from the two stones played at 1. we have either one possible sure win move for P1 (K10) or else at least one possible sure win move for P2. Alrighty, this horse seems to be dead now...

Since this thread is focused on wedge analysis, we'll try eventually to see how probable it is that P2 has a sure win at move 3. If P2 does have a sure win at move 3. this move is most probably the standard 3. ... M8, but it's not impossible that another P2 move could work instead (or as well). And now, back to the main wedge line

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 9, 2019, 6:26 AM

After 7. J8, my current move of choice is 7. ... M8. I'm 99% certain the choice of this move under Boat rules changes at least one sure win P1 8th move in Pente to a sure loss - at the very least, by Dmitri King's and my analysis Up2ng's novelty move 8. L6 appears to lose rather easily in Boat rules while being a sure win in Pente. The proof of this is left as an exercise for the reader.

However, I'm also confident that if a player chooses the most popular move in the database, 8. J6 that I'll be quite happy to be shown the first way that a P1 player finds which beats a reply of 8. ... J9 in Boat - for rating points, bragging rights and $10.

There are at least 20 different 8th move choices P1 has made which appear in the DB, but over 43% of the time, players chose J6, which expert players generally agree is a sure win.

So, when I discovered that Boat rules changed a game between two top players in a fairly popular continuation line after 8. ... J9 from a P1 Pente win to a P1 Boat loss and with no clear idea in my mind how a win could be salvaged, I tried out 8. ... J9 in a tournament game against Dmitri King, and managed to win the game despite him not playing along the hoped for line I initially found (hunt that first line down for yourselves )

After the game analysis between Dmitri and me saw no clear alternatives to try on this line to salvage a P1 Boat win by diverting the line away from a (I suspect) sole sure winning move of 21. ... N14 as the game line went along up to that point.

This game is enough to convince me that I wouldn't want to play 8. L9 as P1 in Boat, despite it being another popular and apparently winning line in Pente.

So far at least, 7. ... M8 looks to be the strongest Boat move for P2, or certainly the strongest I've found to date.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 14, 2019, 6:07 AM

"I'm 99% certain the choice of this move under Boat rules changes at least one sure win P1 8th move in Pente to a sure loss - at the very least, by Dmitri King's and my analysis Up2ng's novelty move 8. L6 appears to lose rather easily in Boat rules while being a sure win in Pente."

I'm going to have to downgrade this assessment - not because I think it is incorrect, but because I can't remember Dmitri King's P2 line and can't seem to find any notes about it either. But in the meantime while we wait for a strong Boat Pente AI to be developed by someone so that all these unresolved Boat questions can be answered, here's an interesting study game on up2ng's novelty wedge line:

Yes, that looks much better for P1. I didn't mean to imply that this was a good P1 line, just an interesting position where P1 could have arrived by mistake

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 16, 2019, 5:29 AM

Wrapping up our exploration of up2ng's novelty P1 wedge variant for now, here are two more games which haijinx and I played using it as a starting point and at the end of the post I'll post the conclusion which Rollie Tesh came to in the analysis thread when analysing play against up2ng as P1 using it.

This turned out to be a very good game. At move 12, I opted for a pair exchange - partly in order to disrupt haijinx's threats, partly to turn up the "Boat rules heat" and partly because if I didn't take the pair at that point it was uncertain that he would let me have it later. In my mind, most games really start becoming Boat games when captured stones reach 6 for one or both of the players. Sometimes Boat rules can make their presence felt before this stage, but in my experience it is comparatively rare for a game to have a Boat ending with a lower capture level, since in those cases it usually involves capturing across fives in order to create or open up a superior threat which must be dealt with by the opponent instead of recreating the 5. A long winded explanation of what I mean by Boat rules heat. Generally, even in Pente games I favor pair exchanges as P2, particularly if I am the player who gets to make the go ahead capture in the sequence. I favor capture exchanges even more in Boat rules. After I make the capture count 8-6 at move 14, if haijinx had taken my pair I would have pair chased for a win in 3 moves. I made my 15th move mostly because I needed to address haijinx's threat, but wanted to show that I could do that in a way which wouldn't be possible in regular Pente. If haijinx had played J7 at move 17, I probably would have played J9. After that; I don't think there is much threat potential on that diagonal for P1. It's possible I might have managed to come up with a 5 in a row win beginning with the move 17. ... O8, but if so it likely would have involved at least one mistake by haijinx with an 8-8 capture count, so I opted for chasing pairs for the win. A great game!

10. ... L9 gave me the K10 stone which made 15. L10 a 4/3 combination threat. Even with Boat rules, those can often be difficult to beat. If at some point after 10. ... L9 had played P9, both M9 and K9 looked worth exploring further as replies.

With a 1. K10 L9 2. N10 wedge opening, the Rollie Tesh and Up2ng Pente moves using this P1 novelty are: 8. M6, K9 9. M7, N7 10.K7, N10 (10. ... M5 rejected due to a forseen P1 forced win) 11.M5, M8 12. O4 - Tesh resigned

All I can recall about the now semi mythical P2 Boat Pente sure win Dmitri King and I thought we had found after 8. M6 was that P2 diverged from the above line at around move 9 or 10, possibly attacking a pair from a different side? and that P1's Pente win took some outside the box thinking to discover and failed under Boat rules.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 16, 2019, 5:58 AM

It's not a wedge game, but I thought this would be a useful example game to illustrate how Boat Pente rules can drastically affect a game position with a 4-4 capture score:

Rainwolf and I were playing this TB game at or near real time game speed, so it shouldn't be considered as a fully analysed TB game. At move 15, I was unhappy with how things looked to be proceeding, so I decided to try L11 instead of making a Pente move of say M14 or M11. I didn't give the move much thought or analysis, it was more a let's see what happens if I do this instead type move. Boat rules heat kicked into high gear, is what happened

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 20, 2019, 6:40 AM

This very tricky just completed game is an excellent example to illustrate why I no longer play the wedge in rated games of Boat Pente as player 1. Players like pente_gon and ivans73 consistently come up with tough moves which I hadn't thought that they would make in Boat games, so they become very tricky to analyse and need repeated reanalysis as the game evolves. This particular wedge line is very rich. I stumbled into my 15th move mostly because it looked like if I let pente_gon play there he seemed to me to have a winning position from that point. So, I looked at some of his possible replies to the move and then made it, hoping that while sacrificing two pairs I would be able to end up with two more pairs myself. The move and position turned out to be much deeper than my initial analysis revealed.

I just figure there have got to be easier P1 Boat wins than the wedge, but will be happy to explore this and other wedge lines further (especially as P2!)

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 29, 2019, 3:09 AM

This is an analysis game on the Pente line Virag used in the first board in this thread, except that haijinx decided to block on the opposite side of the tria. Once I captured his pair, this line resolved into a P2 wedge divergence line which is fairly common in the database - one which seems to be a P1 win. Which just goes to show that when P2 does have a winning line in Pente, it is easy to lose it back in favor of a P1 winning line.

In this game, 17. S12 was designed to set up a stone at R11, since I planned at that point to play 18. M12 and after 18. ... O12 I would play 19. N12 and 20. M14 to set up a double pair threat at P13 on move 21 FTW. Then, after setting up the R11 stone I realized that if haijinx played 18. ... O10 instead of O12 I didn't see any clear continuation for a P1 win.

So, I shifted gears to 18. J14, expecting either M11 or an ignoring move like J12. If J12, then I would have taken at M12 FTW. If M11, then L11 looked good. But, haijinx surprised me with 18. ... M12. I think I could eventually prevail with 19. K12, but then I realized that M11 looked far more promising, provided I protected to L12 after O9. These Boat end games seem to offer a lot more possibilities for defenders than typical Pente games and winning lines can often be hard to spot.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat

Re: Boat Pente wedge analysis
Posted:
Mar 29, 2019, 3:25 AM

The following game boards are from a series of tests pente_gon and I did of the line I played against Dmitri King shown in the third post in this thread. The first one illustrates where pente_gon apparently found a sure win move for P1 in the line. I threw everything I could think of at that move and had no luck cracking it. Let me know if you manage to Ironically, that move 14. loses pretty quickly in regular Pente.

After pente_gon found the excellent 14. P7, I tried to find an earlier divergence point from the line to prevent that move and prevail as P2. The following games illustrate a few of my attempts at this:

As you can see, my attempts were unsuccessful. Perhaps you may be able to improve on them. For now, the wedge may be tentatively safe (though tricky) to play in Boat Pente.

Retired from TB Pente, but still playing live games & exploring variants like D, poof and boat