|
Post by salvor on Apr 19, 2010 15:46:51 GMT -5
Against the double anti weave, it's the even situation of anti versus charm.
It isn't even,it is favourable for S/P.
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 19, 2010 16:19:19 GMT -5
This is the situation I'm referring to: DPPWSCDPW-X PSDWSCPSD-Y & SWDDWSPFPS- PWPPWSSPFPZIt looks pretty even to me. Whatever though. Lets say that situation: DPPWSCD PSDWSCP
SWDDWSP PWPPWSSIs favorable to SP. Will you accept this table? | DPPWSCD PSDWSCP | DPPWSSP PSDWSSS | SWDDWSP PWPPWSS | SP+ | DP+ | SWDDWSS PWPPWSD | DP+ | SP+ |
|
|
|
Post by salvor on Apr 20, 2010 6:43:39 GMT -5
If S/P player moves to SPFPSDW D/P player must counter it, and so the lacks initiative due to pissble dummy PSD(P/S). Where don't forget that either initiative loss is big(D/P prepared two counterspells) or S/P has a chance of making nasty mindgames around the permanency.
Actually I don't thing that left down cell in table is accurate.But other are correct.
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 20, 2010 11:21:49 GMT -5
Ok, now I just need to find out where the bottom left stands, then I can start on my analysis of the cases involving DP shielding on turn 6.
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 20, 2010 17:00:38 GMT -5
12345678901 DPPWSPFPSDW PSDWSDPDWPP DP casts shield instead of anti spell on turn 8 of course. This move is all sorts of crazy! I'm not sure if this worth anything, but I think it might have some potential, and it certainly took FOREVER to find.
Bleh, still didn't get around to analyzing: 12345678 DPPWSCDP PSDWSCPS & SWDDWSSS PWPPWSDS
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 20, 2010 18:27:14 GMT -5
12345678901 DPPWSCPSFWW PSDWSCDPWPP & SWDDWSSSPPC PWPPWSDSWDC DP's advantage. WP opening > CC opening.
123456789 DPPWSCPSF PSDWSCDPW & SWDDWSSSCDP PWPPWSDSC DP's advantage. FW opening > CC opening, and DP will know SP clapped if no spell is cast on turn 9.
1234567890 DPPWSCPSFW PSDWSCDPWP & SWDDWSSSFW PWPPWSDSFW DP's advantage. Shield is cast at the ogre, so it's matter of an ogre versus a goblin. Also DP has better flow.
12345678901 SWDDWSSSPPS PWPPWSDSWDP & DPPWSCPSFWW PSDWSCDPWPW The one below is better, since waving is probably smart even without fear.
1234567890123 SWDDWSSSPPS PWPPWSDSFWP & DPPWSCPSFWW PSDWSCDPWPW This is the most interesting situation. More analysis needed.
|
|
|
Post by salvor on Apr 21, 2010 7:39:53 GMT -5
Don't think that it is true, actually =) At least C/C-player disadvantage is smaller than 2hp (as wp-player could cause light wounds at S/P-player)
DPPWSCPSFWW PSDWSCDPWPW & SWDDWSSSPPC PWPPWSDSWDC
DPPWSCPSFWW PSDWSCDPWPP & SWDDWSSSPPS PWPPWSDSWDP
These variants are missing(there is a decision point on 11th turn, you've analysed just two possible outcomes).
Actually yopu should post where is a "big advantage" and where it is a small advantage. Either estimating this advantage in hp's or other arbitrar scale (you may use my pp-position points if you want).
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 21, 2010 10:38:54 GMT -5
Ach how did I miss that decision point.
I think that brings the situation to about even. Quanifying advantages is beyond my level. For the full analysis I'm not even going to post who has the advantage in a situation, just the resulting situations and perhaps some comments about them.
The comments about who has the advantage, and who doesn't, are so I know where I need to analyze in more detail, and where I can prune the tree. If a player has three options, two you consider disadvantageous and one you consider even, then I can leave the disadvantageous ones out of the analysis.
|
|
|
Post by Dubber on Apr 21, 2010 15:02:42 GMT -5
If a player has three options, two you consider disadvantageous and one you consider even, then I can leave the disadvantageous ones out of the analysis. Respectfully, I must reply: "WRONG!" You need to analyze each path which might be taken by the opponent... or at least each possibly exploitable path... out several turns (say 4 minimum when life points are above 6 or so) Sure, both players' current spellflow may be optimized by your expected path - however, your opponent may not necessarily agree and you need to be ready for the unexpected. See comments about Prioli and his playstyle in these Forums
|
|
scwizard
Ronin Warlock
Lone Wol
Posts: 55
|
Post by scwizard on Apr 21, 2010 16:11:58 GMT -5
Dubber, the fact of the matter is that that is completely infeasible. Every possible path is too large of a problem space to analyze, and there is no way of knowing which subset of each possible path "each path which might be taken by the opponent" or "each possibly exploitable path" is.
That being said, what I can and will do is add the following to my full tree form analysis: Each path that we argue about here, from the point it diverges to the point we stop arguing about it. Each path I encounter in a game, from the point it diverges to the point where one player has a clear advantage. Each path I think up of that I'd like to remember.
Because why not. On the forums though, I'll stick to the problem space I mentioned though, because it's theoretically the most interesting.
|
|
|
Post by salvor on Apr 22, 2010 9:55:11 GMT -5
i haven't seen Prioly's games (only in archive, but them didn't give me anything), but I suppose his style was just close to playing the move which beats the opponents safest move 20% of time and the safest move the rest of time, which was very competeable against typical strategy "play the safest move" which was the most widespread than. Am I right?
|
|
|
Post by Wider, adepte of evil forces on Apr 22, 2010 11:46:37 GMT -5
If a player has three options, two you consider disadvantageous and one you consider even, then I can leave the disadvantageous ones out of the analysis. Respectfully, I must reply: "WRONG!" Depends on the definition of the word "disadvantageous". You can leave out any move A which is dominated by another move B that is: B will always lead to a situation which is better or equal to the situation in which move A would have resulted. "C/-" usually is a bad move and it will be dominated by nearly every other move (There is no way, C/- will leave you in a better position then F/F). But if there is any possible move by your opponent, that will make the seemingy "disadvantageous move" A look better than B, it is not dominated. And it is not easy to say which situations are better than others. Having 15 HP beats 14 HP, but having a giant is not so good when your opponent is about to finish PSDD and sometimes you'd rather be paralysed than not. Wider
|
|
|
Post by Slartucker on Apr 23, 2010 19:05:07 GMT -5
What Wider said. Thanks for posting that Wider.
|
|
|
Post by mikeEB on Apr 23, 2010 21:36:16 GMT -5
What Wider said. Thanks for posting that Wider. Given that dominated cases are omitted, you can also omit moves that are dominated except against the other side's dominated moves. When recursively expanded, this sometimes massively simplifies the analysis, particularly in win-die cases like FoD defenses: (1) continuing the FoD dominates any option that does not have at least 50/50 chance of immediately winning regardless of the opponent's move. (2) The defender can ignore any line that does not involve continuing the FoD and focuses only on lines that possibly defend it. (3) The FoD player's off-hand ignores all defenses that don't try to stop the FoD.
|
|
|
Post by Slartucker on Apr 25, 2010 13:18:36 GMT -5
True, so long as we are strict about our application of the "dominated" label. In this example, there -are- times when it is advantageous to consider non-continuing-FoD moves that do not involve immediately winning. As a simple example, if the FoD gets you a 50-50 chance of winning outright or having a minor advantage, or you can abort the FoD to guarantee a major advantage, I would usually abort the FoD. The only time I would complete it would be if I'd expect lower than 50% odds of winning given the major advantage scenaro -- which is only going to happen if you are playing against a warlock who is a much better player than you.
That doesn't come up all the time, and on many turns you can safely assume that continuing FoD is the best option, but the point is you have to be very, very careful when you eliminate potential moves from an analysis.
|
|