Misere Dodo: I recently added a misere option (opposite object of the game). As hard as Dodo is, Misere Dodo is even harder. You want to remain free and stop your opponent, but to do that, you might have to open channels to free him temporarily. Paradoxical.
Update 1: Nick Bentley's Take On Dodo:
- move straight forward, rather than obliquely forward, especially for stones that start in the middle of your initial formation
- It's better to move up the center of the board and make your opponent flank around you, than the reverse.
If so, it's possible that the dominant strategy is:
- move the pieces at the center of your formation straight forward, starting with the ones at the front of your formation
- then moving the pieces behind them straight forward as well.
- If the other player doesn't adopt this strategy, you win.
- If the other player does adopt this strategy, all the pieces will first gather at the center of the board, at which point there will be cold and fairly obvious a 'who has to move obliquely first' battle, and the outcome of that determines who wins. prob P1.
"This is just an operational hypothesis. Not convinced it's right. Just putting it out there for study."
Another player confirmed that this strategy "absolutely crushes" humans and AI alike. I tried it and it didn't work for me at all. Maybe I'm doing it wrong, but I still can't beat any AIs. Only thing I can think of to explain the discrepancy is that my computer is faster and does more in a second than the tester's computer does in 10 seconds. Hard to imagine though - this crappy Dell I bought in Mongolia 2 years ago is superior to anything. But, if your computer is 12 years old, that could explain it.
The notion of moving up the center of the board and making your opponent "flank around you" is perfectly lovely... assuming your opponent is the accommodating sort and willingly opens up a channel in the center for you. There's no way to force your way up the center, that I'm aware of.
I'm not really fit to comment on strategy, but that being said... The AIs always beat me and they always make plenty of oblique moves. Logically, if you're using Nick's "tight formation" strategy by advancing your front row of checkers directly forward, then nestling your next row in right behind them, etc., you'd do better to hang back (by making oblique moves) and let the enemy come to you. You can keep your formation tighter if you're not on the march. But then if the opposing army approaches with an open channel, your "tight formation" won't be tight for long. I think there is something to it - keeping your army tight, as one of many possible tactics, but it's a long way from a game buster. [Thinking out loud here. Nick stimulated my feeble thought process.] There may be a tactic of rushing some of your checkers forward, trapping a couple of enemy checkers, and later freeing them at the last minute. It's the yin and yang aspect of Dodo. It may be a game of strategy, counter-strategy, counter-counter strategy...
I've tried Nick's strategy several times as Player 1 and Player 2, using different Ludii AIs set at 1 second. I haven't won yet. Maybe the guy who's "absolutely crushing" everything is using a vintage Atari.
Update 2: People are questioning Nick's strategy. 1. "I'm obviously not understanding the strategy, because I am unable to [win]." 2. "The strategy seems to fail against Ai Ai (UCT based AI)."
"The Crusher": "I tried again today and the strategy was much less effective. [...] I think I may have just gotten lucky playing against the AI yesterday." Yes, luck combined with (I believe) running the AI on an antique computer with 10 seconds thinking time. I've never been able to win, with or without Nick's strategy, running AIs at 1/10 second. Of course I'm a dodo of a player, so that has to be considered.
The good news: Nick started a conversation and got me and others thinking about Dodo strategy. Hopefully a talented player will take an interest and clue the rest of us in. I think Dodo is a complex superposition of parity battles - a juxtaposition of interdependent flows - a quicksand beach. Charging ahead and butting horns isn't by itself a strategy. There has to be interplay between charging and stalling. Sometimes you can hold back and wait until your opponent has no choice but to block you. An oblique move only carries your checker half as far as a directly forward move, so a strategic oblique move or two can have a big impact on timing. And it's all about timing. There will be opportunities throughout play to toss a wrench into the gear train and win. Spotting those opportunities is essential. I can spot them - after they happen and it's too late. I gotta turn that around and see them ahead of time.
The consensus on Dodo is that it's hard to know what to do. For most of us, Dodo strategy will be elusive. Hopefully someone will write a strategy guide. One guy beat an AI app, so I know it’s possible to have a clue. I can never beat AI, even set at 1/10 second thinking time. It always stalemates at least one step ahead of me, like magic.
I called it Dodo because the rules are very easy to learn. It didn't occur to me that it would be a hard game to play. Maybe I should have called it Einstein. Someone suggested Cascade, which would have been a good name. It's like two way pachinko.
For the time being, Dodo seems to be a good game for children, and a good game for gifted players. There's not a lot of middle ground right now, though a strategy guide might open that up a little.
There's yin and yang in Dodo. If you rush forward to blockade yourself, then in doing so you also blockade your opponent. If you open a channel for your opponent to pass through, then via a cascading effect, he opens channels for you. This results in a lot of close finishes. It may be unsettling for some, though it’s not a flaw.
I'm a decidedly below average player of games, but after playing Dodo 50 times or so, I feel like I'm starting to develop a vague inkling of what's going on. With familiarity, it becomes easier to see further ahead, and to at least attempt strategy. As with quicksand though, strategies can sink and resurface.
One analytical description of Dodo (by Mike Zapala, the same guy who beat the AI): "Hex4 Dodo is 'bigger' than Checkers in terms of board size, army size, state space, game length, branching factor and game tree complexity. Indeed, its game tree complexity is comparable to International Draughts." It’s a big game on a small board.
Dodo is kind of an essential game. As with all games, the ability to peer ahead into possible futures is important. But without known tactics or strategy, lookahead is all you have. Not to say that Dodo is opaque. It's all there in plain sight right in front of you. It's just a lot to process.
Dodo is a very fair game. Player 1 won 50.5% of 10,000 random playouts. Hex was clocked at 52.45%. Most games have significantly higher than 50.5% random play, move order advantage. Especially opposing army games. I've had to scrap a number of such designs because an overwhelming first move advantage quickly developed.
So there's no need for the pie rule in Dodo (Player 2 having the option of switching colors on his first turn).
I attribute Dodo's fairness to the wide range of possible total moves per game. There can be no passthrough, or complete passthrough, or anything in between.
Dodo isn't vulnerable to mirroring (either by Player 2, or by Player 1 moving to center and then rotationally mirroring Player 2). And it doesn't seem to be vulnerable to any other sort of tricks.
All in all, Dodo is a high quality game, though that won't necessarily translate into popularity. Let's wait and see.
I'm fairly confident that someone much smarter than I will make headway in Dodo and develop concrete tactics and strategy. Stay tuned. I will post any such developments here. Dodo was only born in May, and it's gonna take a little time. It won't be like Reversi where sides are good and corners are even better. It'll be more involved than that.
