Ladders: Set new defaults that fix most testcases and do not slow down Pachi

[pachi.git] / t-regress / games / 2011-06-05-Zen19-pachi2.sgf

(;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Chinese]SZ[19]KM[7.50]TM[1740]OT[10/30 Canadian]
PW[Zen19]PB[pachi2]WR[4d]BR[3d]DT[2011-06-05]GC[185 /S15:bad should just connect at T13
211 T18:bad!{semeai} useless, pachi still thinks the semeai is undecided but white has a lot more liberties. How can the playouts not see this?
217 +S3:bad!{falseeye} pachi must not hope white will end up with a semi-false eye and fill it instead of capturing the cutting stone
219 +S2:alive{falseeye} another verification of the S3 bug above]PC[The KGS Go Server at http://www.gokgs.com/]C[pachi2 [3d\]: GTP Engine for pachi2 (black): Pachi Distributed Engine version 8.99 (Hakugen-devel): I'm playing the distributed engine. When I'm losing, I will resign, if I think I win, I play until you pass. Anyone can send me 'winrate' in private chat to get my assessment of the position.
Zen19 [4d\]: GTP Engine for Zen19 (white): Zen version 7.7
]RE[W+Resign]
;B[oc]BL[1739.323]
;W[pp]WL[1722.852]
;B[pn]BL[1738.474]
;W[np]WL[1714.566]
;B[qk]BL[1726.189]
;W[dd]WL[1697.475]
;B[co]BL[1688.04]
;W[dq]WL[1680.439]
;B[ep]BL[1665.598]
;W[eq]WL[1672.315]
;B[fp]BL[1653.635]
;W[gq]WL[1663.862]
;B[fc]BL[1616.76]
;W[cf]WL[1647.203]
;B[ch]BL[1581.562]
;W[qc]WL[1639.882]
;B[gp]BL[1547.708]
;W[hq]WL[1624.675]
;B[hp]BL[1514.792]
;W[pe]WL[1616.61]
;B[iq]BL[1503.413]
;W[lc]WL[1599.754]
;B[jc]BL[1468.749]
;W[eg]WL[1582.984]
;B[md]BL[1455.95]
;W[ld]WL[1568.011]
;B[le]BL[1444.971]
;W[jd]WL[1551.555]
;B[id]BL[1433.958]
;W[ke]WL[1541.906]
;B[mc]BL[1423.011]
;W[lf]WL[1528.974]
;B[me]BL[1411.757]
;W[lb]WL[1521.223]
;B[kf]BL[1399.469]
;W[jf]WL[1510.346]
;B[kg]BL[1388.745]
;W[jg]WL[1502.247]
;B[kh]BL[1357.129]
;W[jh]WL[1485.818]C[CCY [8k\]: is this a game in a tournament?
]
;B[ki]BL[1346.522]
;W[ic]WL[1478.305]C[NOTA [6k\]: ZvP
ajahuang [6d\]: pachi's opening is good now
]
;B[hc]BL[1319.763]
;W[ib]WL[1463.995]
;B[ie]BL[1309.263]C[Pinguschaf [?\]: j15?
Pinguschaf [?\]: this is gonna be interesting
]
;W[hb]WL[1449.091]
;B[je]BL[1297.716]
;W[kd]WL[1435.253]
;B[hg]BL[1280.962]C[Pinguschaf [?\]: this fight looks better for black
]
;W[ji]WL[1425.213]C[Pinguschaf [?\]: with the thickness waiting on lower side
]
;B[jj]BL[1270.483]
;W[gc]WL[1409.738]
;B[gd]BL[1254.169]
;W[hd]WL[1395.36]C[ajahuang [6d\]: ya black is better
]
;B[he]BL[1243.888]
;W[kj]WL[1387.793]C[DaniPerrin [2k\]: black is fighting well
]
;B[mh]BL[1213.093]C[Pinguschaf [?\]: but im not sure if positional judgement by humans matters at all
]
;W[ij]WL[1379.924]
;B[jk]BL[1202.542]
;W[gh]WL[1371.967]C[dsmic [13k?\]: to bad winrate not working during tournament
]
;B[dg]BL[1171.017]C[jlg [-\]: 53% for black
ajahuang [6d\]: things changed
]
;W[df]WL[1355.899]C[ajahuang [6d\]: L10 was good
dsmic [13k?\]: thx
]
;B[eh]BL[1160.642]
;W[fh]WL[1348.45]C[Bogin [-\]: looks like Zen enters its famous "kill mode" and sets its eye on J15...
]
;B[ef]BL[1131.513]
;W[fg]WL[1334.522]
;B[ik]BL[1119.765]
;W[ei]WL[1320.716]C[ajahuang [6d\]: suddenly pachi is losing
guxxan [6d\]: zEN -> kill mode
DaniPerrin [2k\]: but zen not knows mercy
guxxan [6d\]: cool
]
;B[cj]BL[1088.326]
;W[hh]WL[1313.571]
;B[fe]BL[1059.185]C[Meepy [-\]: h12 didn't look necessary
]
;W[dk]WL[1297.852]C[guxxan [6d\]: yeah, h12 is a waste move
]
;B[gb]BL[1015.543]
;W[hc]WL[1284.494]
;B[cc]BL[979.374]C[DaniPerrin [2k\]: black will live
]
;W[dh]WL[1268.938]C[CCY [8k\]: state of the art tenuki
]
;B[cg]BL[964.824]
;W[dc]WL[1254.751]C[Meepy [-\]: lol
]
;B[eb]BL[950.552]
;W[gf]WL[1243.208]C[guxxan [6d\]: Human does tenuki 
]
;B[ge]BL[936.456]
;W[ga]WL[1227.968]
;B[db]BL[922.721]C[CCY [8k\]: well, im just finding a chance to tell this joke :|
Meepy [-\]: whoops
]
;W[fb]WL[1212.781]
;B[ec]BL[909.462]C[DaniPerrin [2k\]: stupid not to connect
ajahuang [6d\]: oh
]
;W[hk]WL[1197.657]C[ajahuang [6d\]: what is zen doing?
Bogin [-\]: dying :(
]
;B[ii]BL[896.302]C[Meepy [-\]: everything swings around, this is fun :)
]
;W[hj]WL[1182.711]C[jlg [-\]: 57% for black
k81 [4d\]: well - this zen runs on mac so... :)
guxxan [6d\]: oooooooooo
]
;B[ih]BL[883.293]
;W[ck]WL[1175.028]C[Pinguschaf [?\]: maybe zen will win by time :)
]
;B[dj]BL[869.746]C[DaniPerrin [2k\]: zen should resign
]
;W[ej]WL[1160.136]C[CCY [8k\]: winrate?
Bogin [-\]: nah, Zen will kill C12 ;)
]
;B[bi]BL[849.026]C[jlg [-\]: don't understimate the ability of pachi to lose won games
]
;W[il]WL[1153.591]C[ZenAuthor [-\]: 44%
]
;B[jl]BL[836.752]C[CCY [8k\]: thx
]
;W[jm]WL[1139.155]
;B[im]BL[824.962]C[ajahuang [6d\]: Zen was winning to moon
ajahuang [6d\]: now it is losing
]
;W[hl]WL[1125.38]
;B[if]BL[813.419]C[LexC [3k\]: what was the wrong move?
]
;W[kl]WL[1110.935]C[CCY [8k\]: too many state of the art move lol
]
;B[ig]BL[801.95]C[dsmic [13k?\]: f18?
]
;W[kk]WL[1102.327]C[ajahuang [6d\]: D9 was wrong
]
;B[ji]BL[791.102]C[ajahuang [6d\]: should played one more move in the corner
]
;W[bk]WL[1095.355]
;B[lp]BL[760.402]C[Pinguschaf [?\]: hehe
Pinguschaf [?\]: b13
]
;W[ir]WL[1081.697]
;B[jr]BL[730.747]
;W[in]WL[1067.461]
;B[km]BL[720.614]
;W[jn]WL[1053.351]
;B[jp]BL[710.773]
;W[bg]WL[1043.716]C[DaniPerrin [2k\]: zen found b13
]
;B[mn]BL[684.29]
;W[nn]WL[1037.198]C[jlg [-\]: pachi expected b13 next when playing k4
]
;B[mo]BL[674.874]
;W[on]WL[1030.818]
;B[bh]BL[664.895]
;W[bf]WL[1017.719]C[CCY [8k\]: how do you know? are there any logs shown this information?
guxxan [6d\]: I have a question, why bots like to play in one place, then jump to another place, then jump back?
]
;B[pl]BL[639.486]C[me12k: guxxan, because of move "value"
]
;W[lm]WL[1003.871]
;B[mm]BL[628.683]C[guxxan [6d\]: yes, A->B->A
]
;W[qi]WL[991.577]C[guxxan [6d\]: after play B, A value becomes big again?
]
;B[ph]BL[604.8]C[Nickless [-\]: so like
Nickless [-\]: a = 12 points
]
;W[qh]WL[978.931]C[Nickless [-\]: b = 11 points
DaniPerrin [2k\]: because, urgency of these moves alternates globaly
me12k: yes, that's how good bots are made
Nickless [-\]: a2 = 10 points
Nickless [-\]: so it plays a then b then a2
]
;B[qg]BL[583.629]C[guxxan [6d\]: I see
]
;W[pg]WL[966.328]
;B[rg]BL[561.639]
;W[qf]WL[953.978]
;B[og]BL[552.566]C[Bogin [-\]: or could it be that they are similarly big and because of randomness in the playouts it switches back and forth?
DaniPerrin [2k\]: but us, human, we focus too much on local fight
]
;W[pf]WL[941.717]
;B[pi]BL[551.374]
;W[qq]WL[929.951]C[me12k: bogin, not all bots use playouts, but they all associate a "value" to a move
]
;B[ri]BL[530.308]
;W[rf]WL[917.888]C[Bogin [-\]: I see, I was focussing on MC bots
]
;B[kn]BL[510.064]
;W[bq]WL[912.075]C[me12k: MC are the most common, but not all
Bogin [-\]: bots with score-based evaluation functions behave differently of course
]
;B[nr]BL[490.463]
;W[mq]WL[900.848]C[quicksand [1d\]: zen is wrapping up the game
]
;B[fq]BL[471.751]C[DaniPerrin [2k\]: who is winning now ?
]
;W[fr]WL[888.916]
;B[er]BL[464.511]C[CCY [8k\]: there still have points on MC bots. the move with the most playouts will be played
]
;W[gr]WL[877.268]C[quicksand [1d\]: w
me12k: white seems to be winning (but my 12k specticles might be wrong)
CCY [8k\]: the number of playouts is the value of moves
]
;B[lr]BL[446.863]
;W[qn]WL[869.344]
;B[qm]BL[440.96]C[ajahuang [6d\]: pachi played too many small moves
Meepy [-\]: c2 works?
]
;W[rn]WL[857.592]C[guxxan [6d\]: so so
]
;B[cr]BL[425.51]C[me12k: let's see
]
;W[dr]WL[845.541]
;B[cq]BL[424.381]
;W[dp]WL[832.611]C[Nickless [-\]: zen shouldve played c3
]
;B[cp]BL[418.522]C[ajahuang [6d\]: Zen should connect C3
]
;W[do]WL[824.074]
;B[dn]BL[412.613]
;W[cs]WL[818.245]C[Nickless [-\]: it was onlly capturing j2 sente then
]
;B[po]BL[404.317]
;W[qo]WL[805.567]
;B[oq]BL[391.568]C[DaniPerrin [2k\]: zen will cut at c6
]
;W[oo]WL[794.774]
;B[eo]BL[385.759]
;W[es]WL[783.321]
;B[pm]BL[371.955]
;W[br]WL[770.879]C[Meepy [-\]: more back and forth
]
;B[op]BL[357.997]C[guxxan [6d\]: ooo, exchange
guxxan [6d\]: exiciting
]
;W[en]WL[759.065]
;B[pq]BL[348.693]
;W[qp]WL[748.048]C[ajahuang [6d\]: w is good
]
;B[gm]BL[335.269]
;W[em]WL[736.177]
;B[mb]BL[330.112]
;W[ee]WL[724.049]
;B[kb]BL[324.716]
;W[la]WL[717.082]
;B[ma]BL[318.817]C[LexC [3k\]: :)
]
;W[bc]WL[710.286]
;B[cd]BL[303.605]C[Nickless [-\]: zen doesnt see snapback ?
]
;W[bd]WL[698.183]
;B[rh]BL[288.783]C[Pinguschaf [?\]: he doesnt "see" anything
]
;W[gn]WL[686.269]C[Nickless [-\]: oh you smartass
]
;B[sf]BL[275.326]C[Pinguschaf [?\]: :)
Pinguschaf [?\]: but i think he played it because of j19
]
;W[ko]WL[674.295]C[Pinguschaf [?\]: to avoid Ko
]
;B[ln]BL[270.491]
;W[kq]WL[668.802]
;B[jq]BL[256.103]
;W[se]WL[661.466]C[DaniPerrin [2k\]: this game is really confusing
ajahuang [6d\]: strange
me12k: zen might be stronger, but pachi has a nice pic
]
;B[re]BL[242.121]C[Nickless [-\]: im wondering why he played off after n18
]
;W[sg]WL[655.879]C[ajahuang [6d\]: wow pachi
]
;B[bb]BL[229.109]
;W[mr]WL[645.037]
;B[qr]BL[224.751]
;W[rr]WL[635.098]
;B[pr]BL[224.05]C[Imp20 [3k\]: zen gonna win in the end game with moves like s15
]
;W[rs]WL[624.907]
;B[hm]BL[211.128]
;W[hn]WL[618.964]C[Pinguschaf [?\]: maybe s15 is preparing amazing endgame for black :)
]
;B[rc]BL[201.786]
;W[pc]WL[609.27]
;B[qd]BL[195.989]C[Meepy [-\]: guess pachi lost
me12k: doubt it, pachi seems disperate
]
;W[kp]WL[599.018]
;B[lq]BL[191.922]C[ajahuang [6d\]: interesting
]
;W[pd]WL[592.962]C[jlg [-\]: no pachi stil thinks it is winning 62%
]
;B[qb]BL[187.816]
;W[pb]WL[587.694]
;B[rb]BL[186.803]
;W[sd]WL[582.258]
;B[fn]BL[175.017]
;W[fm]WL[574.431]
;B[rd]BL[171.025]C[ajahuang [6d\]: pachi should play T16 first
]
;W[qa]WL[569.25]
;B[ra]BL[170.004]
;W[pa]WL[560.013]
;B[sb]BL[165.451]
;W[kr]WL[554.906]
;B[ks]BL[161.27]C[me12k: pachi seems to loose with 80% acc
]
;W[qe]WL[549.875]
;B[bs]BL[153.781]
;W[as]WL[544.846]
;B[rq]BL[146.855]C[Imp20 [3k\]: what was b1
]
;W[ff]WL[535.255]
;B[ka]BL[135.861]
;W[jb]WL[529.751]C[CCY [8k\]: zen is in cleanup mode
me12k: b1 was: please ignore this move so I can win
]
;B[rp]BL[125.496]C[ajahuang [6d\]: seems pachi misevaluates the top-right group
]
;W[ro]WL[524.61]
;B[sr]BL[114.857]C[DaniPerrin [2k\]: k5 is big
]
;W[sp]WL[519.656]C[me12k: he failed to count libs
]
;B[qs]BL[107.776]
;W[sq]WL[512.927]
;B[rq]BL[106.606]
;W[rp]WL[507.952]C[CCY [8k\]: T18 is temporally not dead, so it maybe wrongly evaluate winrate
]
;B[sm]BL[96.962]
;W[ja]WL[502.583]
;B[kc]BL[90.713]
;W[jo]WL[497.408]C[me12k: w+res ?
]
;B[ha]BL[82.474]
;W[fa]WL[492.245]
;B[el]BL[75.003]
;W[dm]WL[487.181]
;B[dl]BL[68.233]C[ajahuang [6d\]: this game is over
]
;W[cm]WL[480.37]
;B[cl]BL[62.191]
;W[bl]WL[471.868]
;B[be]BL[56.739]
;W[ce]WL[463.689]
;B[ac]BL[51.795]
;W[cb]WL[457.424]
;B[hs]BL[47.063]
;W[is]WL[448.059]
;B[js]BL[43.134]
;W[hr]WL[443.298]C[Imp20 [3k\]: b lost in end game
]
;B[ca]BL[38.204]
;W[ob]WL[438.812]
;B[oe]BL[33.159]
;W[of]WL[428.918]
;B[ls]BL[28.38]
;W[od]WL[424.501]
;B[gl]BL[24.064]
;W[nc]WL[420.147]C[me12k: ,-)
]
;B[fl]BL[19.822]
;W[fo]WL[415.82]
;B[hi]BL[17.742]
;W[gj]WL[411.471]
;B[gi]BL[16.91]
;W[fi]WL[407.105]
;B[bm]BL[12.213]C[jlg [-\]: yes pachi didn't evaluate top right correctly
]
;W[bn]WL[402.688]
;B[gk]BL[7.498]
;W[fj]WL[398.423]
;B[cn]BL[2.49]
;W[am]WL[394.105]
;B[go]BL[29.89]OB[9]
;W[fn]WL[389.853]
;B[nd]BL[27.379]OB[8]
;W[nb]WL[385.616]
;B[na]BL[25.068]OB[7]
;W[nf]WL[381.454]
;B[ne]BL[23.188]OB[6]
;W[mf]WL[377.21]
;B[bj]BL[19.865]OB[5]
;W[di]WL[373.114]
;B[cc]BL[16.473]OB[4]
;W[ed]WL[369.063]
;B[ad]BL[13.179]OB[3]
;W[ae]WL[364.887]
;B[cd]BL[9.436]OB[2]
;W[cb]WL[360.771]
;B[sn]BL[5.37]OB[1]
;W[cc]WL[356.726]
;B[lg]BL[30]OB[10]
;W[fd]WL[352.645]
;B[mg]BL[27.985]OB[9]
;W[da]WL[348.52]
;B[sh]BL[26.196]OB[8]
;W[sf]WL[344.297]
;B[ng]BL[25.365]OB[7]
;W[sc]WL[340.217]C[me12k: no res yet?
jlg [-\]: thanks for the game
ZenAuthor [-\]: thanks for the game
DaniPerrin [2k\]: incredible game
])