.

Thursday, May 5, 2011

IS DOC UP TO THE TASK?

Below this post you will find the current rules of Owlman. At the end of 2010 a change in the rules was made. This was because we discovered a winning strategy for Doc under the original rules. Although my hunch had been that the new ruling might tip the balance in favour of Owlman, this has not yet  been demonstrated on Super Duper Games. Of the last ten games played there, five were won by Doc; five by Owlman.

The basic issue relevant to this rule change is that of repetition of position. This issue, of course, has an impact on many games, Chess and Go, included. Many games (such as Chess) have no problem surviving the rulings necessary. Other games (such as Go) may even be enhanced by these rulings. Nevertheless many modern games, if they are to be played at a high standard of proficiency, are designed so that repetition is avoided altogether.

It has been said many times that if Chess were a newly invented games, no-one would put up with such a demanding set of rules. However, I am not so sure the repetition rule itself is really such a bugbear. After all, if you are playing at a high level, this is the least of your worries. And if you're playing casually, and are in a loop, you would either vary yourself, or offer your opponent a draw. (If he was perverse enough to refuse the draw- but still not vary- then you might do better finding another opponent.)

In Owlman itself,  positions often occur  where Owlman's best move is to attack a particular helper, whereupon  Doc's best option is to move it, whereupon Owlman's best move is to attack it again- and then Doc does best  to move it back where it started! The result of this equilibrium, without any ruling,  would- if neither player wanted to risk a sub optimal move- be a never ending (drawn)  game. This is a problem in Owlman because it happens quite frequently. Formerly it was effectively Owlman who was forced to play sub optimal moves when this occurred- as  Doc could demand that he DIDN'T repeat a position. As explained below, however,  it became clear that this enabled Doc to achieve a long drawn out, and rather boring, win.

The new rules favour Owlman instead. Now there is no rule which prevents repetition, merely a requirement that Doc win before 'nightfall (ie on his 49th move or earlier). The effect is to force Doc to avoid repetition, which means that HE will have to play sub standard moves sometimes (sub standard that is compared  to the proscribed repetition). While this does not seem to allow Owlman to win by use of  a clear cut  strategy, such as can be boiled down to a few principles, nevertheless, it does make the  game very challenging for Doc. If Owlman plays with any verve, Doc's helpers will have run screaming from the woods in next to no time!! Doc, I think, needs to figure out a puzzle- like strategy to inch forward while keeping a sufficient number of advancing helpers on board.

Tricky, but difficult! On the other hand, Doc is still winning games. Either I'm wrong or Owlman isn't swooping as freely (and frighteningly) as he should!

5 comments:

Russ Williams said...

I confess this new rule (time runs out after 50 turns) turned me off... it seems too tedious to count if not playing on a computer. (Do you play it with physical equipment also, or mostly just online?)

I guess such a rule does have the advantage that you could fine-tune the game balance. Too easy for one side? Just add or subtract to the number of turns the game lasts. Possibly even do a "bid to see who plays Doc" kind of thing at the start of the game.

Andrew Perkis said...

The alternative way to discourage Doc from repetition is a rule that allows Owlman to force him to move differently whenever a position recurs. In my opinion this is slightly less awkward for over the board play. But, on the other hand, the 50 move rule works better for online play. Another idea I had was to allow Doc a pre-set number of checkers- say 60- which could be distributed by stacking between the helpers at their normal locations) at the start of play. Each helper loses a checker per move and becomes immobile as a singleton. This game could maybe be fine tuned to work well but I decided it wasn't worth the effort, as too !many 'bits' can also be offputting.
Andrew

Anonymous said...

Alain Dekker here. While I understand the basic strategy Doc was adopting to secure the "long, boring win", I am skeptical that it was proven, ironclad, that Doc was indeed winning.

Repitition in games is a very common problem, though, and dealt with in a myriad of ways. In some games it can be ambiguous. For example, in Chinese Chess many expert players find it quite impossible to explain the repitition rules to new, otherwise intelligent, players!

I think the current rule change is fine. It is true that counting out 49 moves could be tedious, but there is a simple solution. My suggestion is to follow the approach used in Pacru (where you need to get 42 counters onto the board to win). There you count out 42 counters and place the remaining counters away in the box. For Owlman, count out 49 (say counters) before the game starts. Then, whenever Doc moves, Owlman simply takes a counter from Doc. If Doc has not won by the time his last counter is taken...he is lost. The nice thing about this mechanism is that it puts the onus on Owlman to ensure counters get transferred every move, reducing the chances someone will forget on one turn. If Owlman forgets to take a counter before moving himself...thats his own fault!

Andrew Perkis said...

Thanks for that Alain. At the moment I'm exploring new opening ideas for Doc which hopefully do give him a better launch into the game. A small 'chase' game like this is bound to be solvable before too long (just like all the versions of fox and geese for example)but right now I'm thinking that might be a long long way down the line. So it's nice to hear a good suggestion for across the board play.

Andrew Perkis said...

I have reviewed the options re dealing with repetition in this game.
The first two rulings effectively prohibited, first Owlman, and then Doc, from repeating positions. In the first case the game seems open at first but, it turns out, Doc has an unstoppable winning strategy that requires no skill, only adherence to a few principles. Games are then long and boring unless Owlman resigns early.

In the second case (ie with the current 49 move rule) it has looked like Owlman might have a forced win. However,this is not yet clear, and results do no, at this stage, reflect it.

Recently I've looked at two other options: prohibiting players from repeating positions at all, or having no rule on repetition at all. The former is not easily workable, of course. Also, I've discovered, it allows Owlman to win easily. No rule on repetition, on the other hand, would lead to too many drawn games,as situations occur so frequently in which repetition is the best option for both players. Therefore we are left with the challenge- is Doc up to the task? In my last post I indicated he might not be, not because I think it's beginning to look like a closed case, but to throw down a gauntlet. Anyway, I'm taking up the challenge myself- trying to find methods that allow Doc to sort out his opponent. If only those helpers weren't so timid!