(;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Japanese]SZ[19]KM[-5.50]TM[3600]
GN[PaxDeorum-SimbaRules(B) IGS]PW[PaxDeorum]PB[SimbaRules]WR[5d]BR[2k]DT[2018-05-30]CP[
  Copyright (c) PANDANET Inc. 2018
  Permission to reproduce this game is given, provided proper credit is given.
  No warrantee, implied or explicit, is understood.
  Use of this game is an understanding and agreement of this notice.
]EV[Internet Go Server game: PaxDeorum vs SimbaRules]PC[IGS:  igs.joyjoy.net 6969]US[Brought to you by IGS PANDANET]C[
 PaxDeorum 5d : 143 have a nice game
 SimbaRules 2k+: 143 Have a nice game. :)
]RE[W+Resign]LT[]NB[27]NW[33]RR[Normal]
;B[dd]BL[3591]C[White is Carlo, black is Chris.
I (Uberdude / Andrew Simons 4d) watched the game live whilst washing up and doing housework so not reading deeply, more impressions.
File is downloaded from pandanet and has comments from kibitz, but for some reason many ended up on last move instead of when they were said.
Analysis using LeelaZero #145, usually several thousand playouts. Win histogram made with 6k. During the kibitzing I input some of the opening to check its opinion.]
;W[qc]WL[3592]
;B[dp]BL[3585]
;W[pq]WL[3586]
;B[fq]BL[3574]
;W[cc]WL[3574]C[Carlo plays early 3-3 invasion as he has done several times. LZ likes this, 0.11 doesn't. Can't read anything into this.]
;B[cd]BL[3571]
;W[dc]WL[3571]
;B[ec]BL[3570]C[Chris plays hane rather than the f17 jump that is popular and normal bot move these days. -2% says LZ]
;W[eb]WL[3570]
;B[fb]BL[3569]
;W[fc]WL[3569]
;B[ed]BL[3568]
;W[gb]WL[3569]
;B[db]BL[3567]
;W[fa]WL[3568]
;B[cb]BL[3567]
;W[pn]WL[3495]C[This is the move LZ suggests, but adding a move to the 3-4 is obviously the biggest place to spend sente, and this shimari is cool these days, so not suspicious.]
;B[ql]BL[3485]C[The weakness of the high move, but I thought this move betrayed some impatience of Chris, it's already a good exchange for white compared to black getting to approach the corner first. Not sure where else I would play LZ suggests c9 which is a shape now famous that bots like (see AG games too).]
;W[qj]WL[3463]C[
 Uberdude 3d : this should be interesting...

[above comments are from live kibitz\]

Natural pincer, LZ's move but not suspicious.]
(;B[pp]BL[3448]C[The famous tesuji against this big shimari that pros didn't find before but now every man and his dog plays it. LZ prefers p8 jump.]
;W[qp]WL[3400]C[Here white has a few choices, p4 outside and r3 also possible, I would have played this one and LZ too.]
;B[qo]BL[3444]C[joseki...]
;W[po]WL[3396]
;B[op]BL[3443]
;W[qq]WL[3331]
;B[qn]BL[3440]
;W[pm]WL[3330]
;B[pl]BL[3435]C[sometimes black connects r7, LZ agrees this is best.]
(;W[nn]WL[3328]C[a modern joseki against the big high shimari.
After 30k playouts LZ prefers o4 nose]
;B[pd]BL[3315]
;W[pc]WL[3307]C[which side to block is a decision, but both very normal. LZ agrees with this but r16 not much minus.]
;B[nd]BL[3301]C[
 Uberdude 3d : huh, doesn't black need yo defend r7

As can be seen from my live comment, I thought black would defend r7 weakness (e.g. r7, then white likely o4, or s9). Chris is at least a stone stronger than me though, and good at playing flexibly so I thought let's see his plan. FWIW LZ says q16 best at n4.]
;W[od]WL[3282]C[normal joseki, initially LZ says o4 is slightly better, but then likes this (10k), and then likes p10 (around 20k), then back to this (30k). So very normal move, nothing suspicious.]
;B[oe]BL[3299]
;W[oc]WL[3281]
;B[pj]BL[3243]C[This move surprised me, but don't feel strong enough to judge it. LZ wants to play normal human joseki at n14, this -2%]
;W[pi]WL[3273]
;B[oj]BL[3236]
;W[oi]WL[3248]C[LZ says this is a mistake and o15 4% better.]
;B[qi]BL[3235]C[I expected hane at o11 not this cut and LZ agrees it was 6% better.]
;W[qh]WL[3241]
;B[ri]BL[3234]
;W[qm]WL[3147]C[LZ agrees with this, but also quite a natural move]
;B[rm]BL[3199]
;W[rl]WL[3143]
;B[rk]BL[3096]C[it would be nice if black could s6 connect, but I trusted Chris's reading that it didn't work. LZ agrees.]
;W[rn]WL[3137]C[LZ also considers s12 now (more winrate but about half playouts to 40k).]
;B[sl]BL[3094]
;W[ro]WL[3136]C[
 Uberdude 3d : if black defended r7 white would probably have o4 anyway
 Uberdude 3d : cos p3 makes corner suffer

Looked sad to me that p4 no longer cutting stones. Here I checked the position with LZ, W was about 70%.]
;B[rh]BL[3019]C[
 Uberdude 3d : my guess is LZ says w 80%

I did wonder about black here live as it's painful to let white get it in sente and often a strong reverse sente move, but was surprised Chris played it as it seems even more painful to be slow/inefficient with the ponnuki adjacent. LZ suggests m15.
]
;W[gp]WL[3081]C[
 Uberdude 3d : ah no,. only 17 after trade

LZ #145 does not consider this move, but o15 cut (natural to me) or c9 (black's good place is mine). Maybe an old network does as some anonymous person online claimed Carlo used in this game. ]
;B[iq]BL[2908]C[Tenuki a shoulder hit used to be pretty crazy, but this shape is something we can learn from bots these days. LZ's #1 move, well done Chris!]
;W[gq]WL[3069]
;B[fp]BL[2907]
;W[go]WL[3067]C[
 Uberdude 3d : 70

Natural move to my eyes, but LZ prefers f2 by 4%.]
;B[kp]BL[2903]C[This or m3 or j5? My 1d wife thought j5 was much better, I wasn't sure. LZ says j5 best. I can understand this plan too though to make the better base inside that with 2-space extension.]
;W[io]WL[3059]C[
 PaxDeorum 5d : 143 ops :(
 SimbaRules 2k+: 143 Do the league rules permit takebacks?
 PaxDeorum 5d : 143 my mouse slipped :(
 PaxDeorum 5d : 143 i don't know
 SimbaRules 2k+: 143 I'm happy to let you have it back if it's okay with the organisers.
 PaxDeorum 5d : 143 it was j5
 SimbaRules 2k+: 143 Yes, I figured.
 Jad 9k : misclick?
 Uberdude 3d : almost certainly
 Uberdude 3d : wanted j5 i guess
 PaxDeorum 5d : 143 i ask to the organizer
 xyper 1k+: undo possible?
 SimbaRules 2k+: 143 Okay.
 Uberdude 3d : technically, depends on client
 Uberdude 3d : i don't know league rules on undo
 PaxDeorum 5d : 143 ok, they say that it is up to you
 Fede 1k : if Simba agrees, undo can be done
 Fede 1k : if Simba's client has the undo function or the console
 Uberdude 3d : L group is dead!
 Fede 1k : (else players could agree on a move for Black too)
 Jad 9k : lol
 Pierone 2d+: hope not!
 xyper 1k+: Lets see how it goes...
 SimbaRules 2k+: 143 Yeah, I don't want to win/lose based on something like that.

[Carlo played h5 misclick\]

LZ starts liking o15 or c7, then settles on k5 not j5 (which only gets a few playouts). This move didn't surprise me though, if Carlo had played k5 that would seem more suspicious as beyond my 4d intuition.
]
;B[mq]BL[2901]C[makes the base]
;W[kn]WL[3026]C[I thought white would tenuki, e.g. I like c4 or something around c6. LZ likes c7. Doesn't consider this move]
;B[dm]BL[2652]C[This is the move I expected form Chris as a common shape to see these days, assuming bottom group is fine. LZ prefers c8.]
(;W[fd]WL[2919]C[One of my biggest surprises this game was neither player played at the bottom for ages, I thought black might die and L6 was maybe a bit slow if it wasn't sente. So I thought black might defend at p3 if he got a chance (but maybe Chris felt behind so no time to do it) or white attack with o2. See var for my "let's play obvious moves" first reading.

This push is barely considered by LZ, Carlo seems to like it, he also played it early vs Dragos. LZ likes c9 or o2 or o15. -2.5%]
;B[ne]BL[2372]C[So Chris got to connect, as I expected (not sure exactly which connection, LZ starts liking n15 but this #1 after 15k).]
;W[qf]WL[2872]C[LZ initially wants to attack at o2, as I worried about during game, but after 10k likes this which is the obvious local move.]
;B[qd]BL[2274]C[I thought Chris might defend at p3 now, but instead he went for this sharp move I didn't even consider (LZ neither). This actually reminded me of his game against someone else from Italy last year, he is good at sharp local fighting like this. The idea seems to be if white s16 then s15 etc and black gets o17 in sente which is nice.]
;W[pe]WL[2766]C[Both players next few moves proceed same as LZ, make sense.]
;B[rd]BL[2272]
;W[rc]WL[2763]C[Worth reading other moves here, white can probably capture black but gets squeezed and o17 sente.]
;B[se]BL[2271]
;W[mc]WL[2751]C[LZ consider this a good time to o2 instead. This move seems pretty normal too and is considered a bit (-2%).]
;B[ni]BL[2261]C[Natural move. Live I wondered if white had to hane or had a good dodge, in which case maybe black p13, white here then q12 was an ok compromise. LZ agrees this is only move.]
;W[nh]WL[2748]
;B[oh]BL[2218]
(;W[ph]WL[2738]C[This is a crucial point in the game. Watching live I thought connect was a mistake and Carlo should p14, if black takes then white o14 and is happy to minimize loss and attack o15 stones. So black probably doesn't take but then something more complicated happens. I added a variation of LeelaZero living inside using the thinness of the monkey jump connnection: if Carlo had played this I would have found it suspicious as it's a very neat sequence. Connecting an atari is a very natural/human thing to do.]
;B[og]BL[2216]
;W[mi]WL[2651]C[So Carlo dies, and chooses to make some sacrifice exchanges. LZ would attack the lower group (h2 not o2 now) and gives white 48%]
;B[nj]BL[2214]
;W[ng]WL[2650]
;B[pf]BL[2197]
;W[qe]WL[2636]
;B[qg]BL[2135]
;W[le]WL[2626]C[At this point LZ would o2 with 30% winrate (so these exchanges quite bad). Watching the game live I was feeling happy for Chris and thought he had recovered from the earlier mistakes (mostly losing r5 and s12) and the game was probably about even now without counting it. (You can see my kibitz  erroneously on the last move of the game " Uberdude 3d : assuming bottom doesn't die i think b recovered"). This was a perfect time for the p3 defence which I had wanted for a while. LZ agrees with me and gives p3 as best move with 75% win. " Uberdude 3d : so is p3 too slow, or d10 or f15?"]
;B[nc]BL[2002]C[But my heart sank when I saw this move. Losing a liberty is pretty bad, just n16 directly if you want to (LZ likes that too, actually prefers to make n16 m17 exchange and then p3 after 10k playouts). But why even play in this area? some reduction around k15 would seem a calmer and more collected move if you don't want to play on lower side or left.]
;W[nb]WL[2574]
;B[md]BL[2000]
;W[lc]WL[2573]
;B[ld]BL[1999]
;W[kd]WL[2572]
;B[ke]BL[1999]C[This sequence seemed impatient. I guess Chris was hoping white would k16 now and he could m14. ]
;W[mf]WL[2476]C[However, Carlo played this excellent move (Chris praised it in his game report). Now we see the badness of the liberty-losing o17 push which set my spider sense tingling. Watching live I didn't see this move but did see white n15 was sente for the liberty shortage if he wanted to save the cutting stone in sente and fight. LZ does find this kosumi. So Carlo playing it I would say is slightly suspicious (and lots of little bits of suspicion can accumulate) but not massively so. Perhaps we could pose this as a problem on the forum like I did for that reduction move vs Kulkov.]
(;B[pg]BL[1904]C[I wondered if p14 was better (it make n15 not sente which affects the m13 peep later), LZ agrees its better, but says interesting n12 is best and just take sente to tenuki]
;W[kf]WL[2469]C[LZ would rather play e15 initially or then f2 or o2 and we have the lower side life and death problem. Continuing to fight here using the n14 kosumi seems natural too though to me.]
;B[je]BL[1856]
;W[jf]WL[2468]
;B[ie]BL[1855]
;W[jc]WL[2466]
;B[if]BL[1849]C[Frmom L14 LZ agrees with B and W. Watching the game I felt bad about black running this group as white's centre stones are not so weak]
;W[ig]WL[2465]C[Here LZ says j12 jump is better, hane seems the more obvious move to me as a 4d human.]
;B[mb]BL[1744]C[I didn't really get the point of this, is there some slight endgame benefit in corner later?]
;W[lb]WL[2438]
;B[hg]BL[1742]C[I seem to recall Chris took quite some time around here. My thinking was if g14 jump you can't get cut, but white just nicey forces black to save the junk stones. So with hane black is probably happy if white cuts because you can sacrifice and build the left]
(;W[hh]WL[2436]C[Hane or extend? 2 obvious moves, Carlo does hane, LZ slightly prefers extend]
;B[ih]BL[1650]
;W[jg]WL[2421]
;B[gh]BL[1648]C[Looks like Chris is indeed happy to sacrifice, so by offering that he can get his head out a bit more vs extend.]
;W[hi]WL[2310]C[Carlo doesn't take the bait and give a ponnuki]
;B[gg]BL[1646]
;W[gi]WL[2257]C[Here LZ wants to attack at o2, the move I had been worrying about for ages. Instead we get a ddk style pushy pushy battle across the board.]
;B[fi]BL[1599]
;W[fj]WL[2254]C[LZ thinks now was a good time for b17.]
;B[ei]BL[1597]C[LZ prefers hane]
;W[ej]WL[2251]
;B[dj]BL[1594]
;W[bc]WL[2205]
;B[bb]BL[1579]
;W[dk]WL[1829]
(;B[cj]BL[1576]C[I didn't like this move, I thought/kibitzed Chris should have connected, if white goes down then cut. Maybe white can live in left side but if B can get some centre influence k6 cut looks plausible. LZ agrees this was a significant (9%) mistake.]
;W[ck]WL[1811]
(;B[ek]BL[1458]
;W[gk]WL[1700]
;B[bk]BL[1433]
;W[el]WL[1697]
;B[cl]BL[925]
;W[dl]WL[1695]
;B[em]BL[924]
;W[fl]WL[1693]
;B[bm]BL[923]C[So the whole sequence of running with the L15 stones ended up rather a failure as the white centre group was never weak, black didn't gain much on left compared to just spending a move there and h14 f12 cuts still bad aji with b17 stones. But I don't think Carlo needed to play any particularly good moves, it was all a very natural hane/nobi pushing battle. Maybe at some point Chris could have just cut his losses and sacrificed the L15 stones (LZ says no, e.g. would indeed pull out with j14 rather than sacirifice), but during a game it's easy to get tunnel vision whereas kibitzers can be more emotionally detached.]
;W[ff]WL[1585]C[Normal move, LZ finds e13 as even better (would give white 76% vs 73%, the black lead after top right fighting has switched to white now)..]
;B[eg]BL[839]C[The obvious defence, LZ agrees.]
;W[df]WL[1584]C[This was a nice move I didn't see, LZ does but then says o2 (again!) is better. I'm sure plenty of 5 or 6 dans who are concentrating on their game rather than doing the washing up as I was could find it though. So only a little suspicious, like n14.]
;B[bd]BL[749]C[Interesting LZ says black should p3 now sand sacrifice the f11 stones if white d11 cut.]
;W[ef]WL[1536]C[Again LZ says o2. I was really surprised neither player thought that area important for so long. But continuing here does seem human.]
;B[ge]BL[721]
;W[fe]WL[1515]
;B[dg]BL[687]C[Chris defends as LZ would.]
;W[fr]WL[1510]C[Finally we get play here, LZ would o2.]
;B[oq]BL[620]C[So was this Chris's plan all along, if white starts on left then p3, and if white starts at o2 then h2? LZ says B h2 before p3.]
;W[hr]WL[1421]C[LZ thinks d2 is significantly better than this because black can answer at e2 and not die.]
;B[ir]BL[572]
;W[dq]WL[1414]C[LZ's move, but pretty obvious after the h2 defence.]
;B[cq]BL[560]
;W[dr]WL[1411]
;B[hc]BL[490]
;W[hd]WL[1150]C[A sharp counrer which I also saw live.]
;B[lg]BL[429]
;W[lf]WL[1125]
;B[mj]BL[240]
;W[cp]WL[1047]C[An obvious place as black didn't defend, LZ prefers g16.]
;B[gd]BL[213]C[LZ would play here too.]
;W[gc]WL[1044]C[LZ would not cut bit m11, but this is the obvious local move. Humans to tend to not tenuki as much as they should.]
(;B[he]BL[202]C[Instead of this LZ sees some chances for black at j11, make use of n10. I think a 5/6d who was not feeling emtionally crushed (as I suspect Chris was) and had more experience of reversing losing games against strong players under tournament pressure could find this move as a way to fight back (e.g. Andrew Kay as black could probably beat me as white).]
;W[gf]WL[1044]C[LZ would m11.]
;B[hf]BL[200]
(;W[lj]WL[820]C[LZ would m11 extend after 5k, then prefers l10 jump. That is a rather advanced shape to my eyes, so if played I would have found suspicious.]
;B[hb]BL[168]
;W[ib]WL[818]C[So now we have the famous move from the anonymous "CarloCheating" account (Leela 0.11 doesn't see this j18 but just h19, both let white connect but this is better, LZ sees this).  However, I think it's not unsurprising for a 4d+ to see this (I did), also if Carlo thought he needed to defend he could have done the previous move isntead of m10, though then j11 might have made some problems in the centre.]
;B[jd]BL[498]
;W[ha]WL[757]
;B[kc]BL[495]
;W[kb]WL[754]C[
Black could try m11 or j11 to see if white messes up but just resigned now. Kibitz of final and other moves from game below

 PaxDeorum 5d : 143 thanks
 SimbaRules 2k+: 143 np ^^
 xyper 1k+: mmhhhhh
 Uberdude 3d : i think p3 is needed
 Uberdude 3d : o15 looks big
 Uberdude 3d : assumign they both think b is alive after o2 which i'm not conviced is true
 Uberdude 3d : maybe b can get m5 in sente to help
 Uberdude 3d : but its not really sente
 Uberdude 3d : tryingf to make o17 sente
 Uberdude 3d : q12 seems mistake
 Uberdude 3d : assuming bottom doesn't die i think b recovered
 Uberdude 3d : so is p3 too slow, or d10 or f15?
 Uberdude 3d : not convineced this push is good
 Uberdude 3d : the liberty is important
 Uberdude 3d : could ave n16 directly
 Uberdude 3d : if you erally tihnk this is important area
 Uberdude 3d : as white n15 sente if he wants to make a mess
 Uberdude 3d : was p14 better?
 Uberdude 3d : hane
 Uberdude 3d : if w cuts then sacrifice gladly
 Uberdude 3d : b spoilt his lead just as he'd got it
 Uberdude 3d : seems like b should have connected
 Uberdude 3d : all aboard the nobi train!
 Uberdude 3d : if c10 was connected and then w go down and b cut maybe b could let w live in left and cut k6
 Uberdude 3d : d14 very nice
 Uberdude 3d : p3 ?
 Uberdude 3d : or is d11 problem
 afar 2d+: p3lease
 Uberdude 3d : wowzers
 Fede 1k : looks like your prayers were heard
 Uberdude 3d : i was opeing b played it when he had sente =)
 Uberdude 3d : i suppose he plans to tenuki d2 to reduce middle?
 Uberdude 3d : but w should win like that
 Uberdude 3d : this move has problem
 Uberdude 3d : h16 sente for h14
 Uberdude 3d : things are not looing good for our hero
]
;OS[Andrius][Andro921][Coorchac][Fede][Ginkko][Harleqin][JRDK][Jad][JeffChang][Kiyomiya][Lukan][Madsen99][Murseppas][Nighthawk][Pierone][Saeki][Uberdude][Wade][Wiles][Wolf94][Ziridan][afar][dsatkas][eagle888][erdem][go2win][jaqwith][k22452245k][kisekigo][ld22][mitu3634][noviaprile][observer99][pizza][taxxcell][tokiwa0728][tyam][xyper][yongo2])
(;W[ic]))
(;B[ii]
;W[lj]
;B[ij]
;W[hk]
;B[lk]
;W[kj]
;B[kk]
;W[ki]
;B[ik]
;W[lh]))
(;B[gj]C[How about this cut? My 1d wife suggested it and LZ agrees 3% better.]))
(;B[di]
;W[ck]
(;B[ek])
(;B[gj])))
(;W[hf]
;B[ih]
;W[jg]
;B[gg]))
(;B[mh]
;W[of]
;B[nf]
;W[mg]
;B[rf]
;W[lh]
;B[je]
;W[jd]
;B[gm]))
(;W[of]
;B[nf]
;W[og]
;B[ng]
;W[ph]
;B[mh]
;W[rg]
;B[sg]
;W[re]
(;B[sf]
;W[qe]
;B[sd]
;W[sc]
;B[rj]
;W[rf]
;B[sh]
;W[pg])
(;B[oh]C[LZ thinks this better and we have a ko])))
(;W[nr]
(;B[mr]
;W[hr]
(;B[ir]
;W[fr]
;B[er]
;W[jp]
;B[jq]
;W[lp]
(;B[kq]
;W[ko]C[This kind of thing is locally dead. Some aji but this is the basic shape white aims at.])
(;B[ko]C[So need to read things like this.]))
(;B[ln]C[does black do sometihng like this?])
(;B[mo]))
(;B[hr]C[You can also do self-damaging moves like this in an emergency.])))
(;W[np]))
(;B[ol]))