Sente:

 Cyril

Gote: 

 PSProgrammer 

@Sente: Reasonable, but lacking the offensive potential of Rs*15L or similar.

@Gote: Restoring sov to both tiles; Gote takes control of the central garden, but with little effect; He also falls behind in board control as he doesn't add another tile. 

@Sente: Looking to counteract potential meld plans. Since no meld is in sight, maybe a better idea would have been to take 2j instead, as it is a weak point in Sente's position so far. 

@Gote: Clearly needed if going for a meld, but questionable since it uses a whole move.

@Sente: Restoring neutral garden sov to prevent Gote's Ly*14n for 4 points. It remains unclear if Gote should go that way, since he would suffer from loosing his board influence.

@Gote: Exploiting Sente's over-aggression and occupying an important strategical position. Consider Ly 2j-3j, eliminating Sente's sov in the upper and lower left area simultaneously.

@Sente: The threat of Jd 15h-13h with Js 14f-13h remains easily dealt with by blocking on 14g, but Gote must do so now, since the garden is off limits when Sente shifts the Jade over, creating sov on the Jasmin.

@Gote: Very unconventional and surprising; Gote doesn't deal with the meld threat around 13h, but threatens his own with a potential Jd*7n for a Jd-Rd-Ly meld.

@Sente: Too concerned about Gote's plans. Jd 15h-12h would have been better, since Sente can go for a meld exchange while winning one turn after Gote's Jd*7n, for example with Ly 10r-7o. This forces Gote to complete the meld immediately and earns Sente the initiative. With Jd 6n-9n, both melds are prevented temporarily.

@Gote: Now that Gote has the initiative, he can prevent Sente's formerly unavoidable meld, or go for 16k, threatening the Ly and opening up the upper right area for a strong Lily drop on 14n.

[End of session one]

Cyril: 

Since it was my turn, I used the time during our break after completing the first session to consider my options in this position. I needed to prevent Gote from finishing the 10n-11n meld, but he could complete it with a Jade on either 10p or 10L, which both needed to be dealt with; Gote had given the initiative back to me with C 5o-6o, so I had multiple ideas to explore:

Jd*8p, covering the upper spot while threatening the Ly; should Gote ignore it and go for the lower option with Jd*5L; I could do C*7n next and block the way on 9L before he gets there. However, the idea uses Jade and C tiles, both of which I had very few left, so I disliked it.

Another option would be Js*12p and then moving it to 11o, but Gote wins with Jd*11p, blocking the way and getting there in time.

I also considered Js*11q followed by Rs*8p; the threats on Gote's Js/C would have bought some time, while I get an opportunity to set up a stable block with Js 11q-10o. However, the lower passage is completely undealt with. Also, the two drops don't look like they would really force Gote to react. He could let the 12o Js be captured (followed by the 14n Lily, by the way) and it would still be a 3-vs-2 exchange for him. But the resulting board position is in Sente's favor, which must be considered as well; The whole variation seemed unclear to me and I decided against it.

The move that I ended up playing is Rs*8q, a very passive way to block off that upper meld constellation. If Gote decided to go for the lower one instead, I could still go C*7n if needed.

Cyril: I actually failed to see that PSProg threatened a meld here (with Ly 7o-10L), so Sente's move is not on-point here. Threatening the 14n Lily is a good idea, but Jd 9n-8n would have been better.

@Gote: Gote needs to complete that meld threat on 10L immediately. This move probably doesn't have enough offensive potential to secure the meld on its own (if Sente had been aware of it, of course).

@Sente: Note that this Lily is at zero hmp and still protected. My idea was to give an obvious threat against the 14n Lily, while secretly aiming at 10L to block Gote's meld after all. Again, Jd 9n-8n would have been more efficient and ressource-friendly.

@Gote: Very unexpected, to say the least. Sente should use the time wisely and block 8n immediately.

Ly 7o-10L

ring[10L-10n-11n]

Js*7n !?

Cyril: I contemplated this move by PSProg rather thoroughly, but even now after having replayed the game several times, I don't yet understand the meaning of that Js ;-)

@Sente: Even if Gote goes for Ly 12m-10o now, Sente can still secure the meld with Rd*6m, so he doesn't finish it right away, but threatens the C first. Notice that the Rs can also shift over to 13k later and expand Sente's harmony chain in that area if needed.

Rd 12o-9o

ring[9o-9n-10n]

Cyril: Not very clear play by me here. Looking back on it, even to myself the purpose of this move remains unknown.

@Sente: Looks redundant and unnecessary, expecially considering Sente's heavy Chrys use so far. Prepares a meld around 5j-6j, but a Js*5L would have been better.

@Gote: Unexpected, and probably not as good as shifting the Chrys over to the same spot. Now the 7n C is out of position and doesn't benefit Gote's position.

Ly 15g-13i

ring[13i-12j-12h]

@Sente: A reasonable idea, but again, Js 16m-14n would have been the way to go.

Rd*14n !?

Cyril: That was a very risky sacrifice. I should probably have chosen the more stable Jd*12o, but I disliked it because I had only two Jades left at the time. The consequences of this Rd drop are very unclear and can easily go either way. Seeing as PSProg was at the advantage in this area anyways (consider that Jd 10nx10L can finish a secondary meld with the 7k Lily even if I destroy the 10L-12m-13L one), I hoped that the chaos resulting from a daring move like this would even the position through capture exchanges.

@Gote: A very strong defensive move utilizing the 15L Rose to resolve the situation and nullify Sente's aggression. Gote's 14m Rd that was the target of Sente's sacrifice is now protected, while Sente's 14n Rd is in still in danger, effectively turning Sente's aggression back on himself.

Cyril: I was very impressed when PSProg played this.

@Sente: Notice that this saves the 14n Rd and threatens the 16n Js. However, the 13L C is not yet in disharmony.

@Sente: That Lily comes too late to contest Gote's meld, but threatens to capture the 14m Rd, which would bring Sente's 14n Rd into a good position to create his own meld in the area.

Ly 15L-12i

@Sente: Makes good use of the Ly and removes a disharony from 13n, but the meld is easily blocked off.

41

Js 9o-11n

ring[11n-12L-10k-12m-13m]

(Diagram 4) 

 @Sente: Questionable. It was Sente's last C, but it secures a meld on 15k. Maybe Ly 13n-12n would have been a better idea.

@Gote: A very impressive and entirely unexpected meld by Gote that catapults him ahead of Sente in points for the first time in the game. Only Gotes big drop in board influence and Sente's resulting ability to secure at least three more points for himself make the move seem questionable. Notice however that Gote clearly has a much higher tile efficiency in terms of turning dropped tiles into actual melds, and - unlike Sente - doesn't suffer from an increasingly empty tile reserve.

C 16k-15k

ring[15k-14L-14j]

@Sente: Sente evades the threat on his Rhod by casually completing the meld, still maintaining considerable control over the area with his 12i, 12n and 13n tiles.

@Gote: Improves the Jade's position, but doesn't pose a threat and therefore gives the initiative away to Sente. 12g is protected by the Js, but 14i would have been a good idea.

@Sente: Hoping for a meld with Rd 12n-12L and Js 15k-13j.

@Sente; Sente is too easily distracted by Gote's maneuver. Trying to attack the Ly (that is now at -1 anyways, due to Gote's new Jade) is a neat idea, but not practical. A better idea would have been Js 15k-13j, which secures the meld even if Gote blocks off the Rd (consider Rd*16i).

@Gote: Cleverly makes the Ly less attractive by threatening a recapture on the 13n Ly. Sente could go for the exchange, but doesn't have to, since Gote's move only works when he does.

@Gote: Allows Sente to stabilize his position and protect the Ly; going for the straight capture on 13n would have been a better idea.

@Gote: A nice move that threatens the Js while adding tiles for potential meld in the area.

C*15p

@Sente: Probably a mistake. The idea was to protect the Js while secretly setting up a discovery threat on the Rs behind it, but Sente only has two Lilies left and this kind of move is not enough to legitimize using one of them.

@Sente: The 14o Js must stay in place to protect the Rd; Sente removes the Rd to free his Js as the situation becomes increasingly dangerous for it to remain there.

@Gote: A small move that opens up a capture on the 9n Rd, making the coming exchange even or maybe advantageous for Gote due to the resulting position.

Rd 6i-6k 

ring[6k-5k-5j]

Ly 9k-10L 

ring[10L-9m-11n]