Bet On Cobalt Lets You Gamble On Bot Matches

Cobalt is a sidescrolling shooter that’s getting a finished release on February 2nd, and between now and then its developers are running Bet On Cobalt. It’s two weeks of non-stop Cobalt matches between teams of bots, around which the Twitch audience can gamble with fake money on who they think will win.

The bots are playing a mode called Team Strike. Viewers are given 100 ‘volts’ on which to place to bets as to which team they think will win, by typing “ibet blue 50” or “!bet blue 50” and so on into the Twitch text chat. You can also go further by betting on particular achievements happening within a round, such as either team scoring a combo or a single bot taking out the entire opposing team. Bet well and you’ll win more volts with which to bet more.

Or with which to buy things. Although volts have no real money value, they can be used to modify your profile within Bet On Cobalt, to name a bot, make a bot speak, or to trigger “special events”. Then, if you’re really successful, you’re in with a chance of winning “real-life” prizes on February 2nd at 12pm CET. There’s more details about all of this, including more on how to bet and how to win, over on the Mojang site.

Twitch phenomena seem to rise and fall in the time it takes Bob Ross to paint a friendly tree, but the one I still watch is Salty Bet. That stream pits bot-controlled characters against one another in heavily moddable fighting game engine M.U.G.E.N. It’s funny, owing to the odd match-ups of anime girls, Street Fighter characters and, to pick an example on my screen right now, Dr. Robotnik from Sonic, as well as being oddly compelling because it’s rarely obvious who will win. It’s the most likely inspiration for Bet On Cobalt, although I’m not sure that Cobalt – with its teams of more or less identical little robots – has the same lasting appeal to me as a spectator stream.

I do like Cobalt a lot, though. It’s been a long time since I played it, but there’s a lot of finesse to it; I especially like how near-misses trigger slow motion and how rolling allows your character to deflect bullets. It made for some shouty, dramatic matches when I played it in local multiplayer a few years back.


  1. Zankman says:

    Cobalt looks very interesting, in terms of visuals and gameplay it is the kind of game that just draws me in.

    As for this Stream, I think that this is a very ingenious idea and I give props to Mojang for it.

    I do agree, though; I reckon Salty Bet has more lasting appeal, what with the various different and colorful characters, as well as with the fact that viewers can follow the “careers” of these characters (and lie to newcomers about who to all-in on).

    What I don’t like about Bet On Cobalt from what I see is that betting happens before each round; Why not have the bet take place before each match?

    This way bets are more frequent I guess, but, it would be much more compelling to bet on the outcome of an entire match and then sit back and enjoy the show (and often hope for comebacks).

    • Kitsunin says:

      I’m not really sure how they should change it, but yes, the rounds are way too short for the stream to actually be particularly entertaining.

      • Isogash says:

        I’m not sure I agree with either of you about match betting. If you bet on a match, you might as well just flip a coin because the teams both start off random and balanced. What makes the Cobalt betting interesting is that the AI bots upgrade themselves each turn, so you can predict which team will win based on how strong their bots are.

  2. Baines says:

    I’ve tried watching Salty Bet in the past, but it is simply boring. It suffers the balance problem of using MUGEN, in that character balance is completely non-existent. It isn’t just that different people made characters to completely different styles and standards, but that many are also intentionally broken.

    So every match that I ever saw on Salty Bet was completely one-sided, either decided by which character landed the first hit (when two broken characters faced off) or decided when the characters were originally selected (when a more broken character faces a less broken one).

    I wished that they’d take the time to at least try to rebalance the characters that they use, but that would take a lot of effort and would probably also cause some creators to throw hissy fits. (I used to be involved with the MUGEN scene, so I know how fragile some of the people involved could be when they felt someone else was using or modifying their work without permission. And yes, it was lost on them that that was exactly what they were doing with their sprite ripped “creations”.)

    Perhaps more interesting would be if they actually had real people playing the games (even if they weren’t pros), a never-ending stream of fights, with various performance stats recorded.

    Regardless, I’m not sure Cobalt will generate enough interest to last more than a short period. Of course, if it is only running two weeks, then that doesn’t really matter.

    • ButteringSundays says:

      In the betting world ‘character balance’ would be rather boring, giving equal odds to all parties.

      Did you complain that Mohamed Ali was OP?

      • Baines says:

        Fighting games don’t achieve pure balance. And unless you just like to always see first round knock-outs, matches between more balanced opponents can be more interesting to watch.

        At its best case, combining random MUGEN characters is like having boxers, wrestlers, basketball players, badminton players, and ultimate frisbee players all compete against each other in UFC matches. The characters, when made to any standards, are made to wildly different standards. More normally, it is more extreme that that, and closer to having adults fight robots, children, chihuahuas, and tanks.

        There isn’t even much time for chat, not when the pre-match timer lasts longer than the match itself, and the winner is decided within the first two seconds no matter how long the match runs anyway.

    • vlonk says:

      I once watched through a saltybet tournament and realized that “broken” characters are not a problem at all. It is actually part of the maindraw for an aspiring newcomer in the betting. Sure a lot of fights where onesided and in the first round – without any prior knowledge about the characters – betting was pure luck. But that changed in the following rounds when the weaker fighters where filtered out. It took imagination and skill to understand what “broken” skillset will trump whom. Because the bets are a zero sum game with the other ppl in the chat you can actually interact with them, the whole twitch chat, in a meaningful way. You are all in on the guessing game. It’s a powerful mechanic and MUGEN lends itself very well to this. Also with some bots there is a “wild card” excitement going on: Will the bot use his zero-frame instant teleportskill of death? Will the super regeneration of the other bot win out? Plenty of potential for comebacks and unexpected outcomes because of the insanely big roster of fighters.

      • SuddenSight says:

        I agree with this. It isn’t really interesting to watch computers play fighting games, but it is interesting to try and guess what characters do based on very little information.

        The fact that its computers playing also makes it fairly interesting. A lot of upper tier characters have perfect loops, but if the computer doesn’t use it it doesn’t matter.

        So you are trying to guess how good a character’s moveset is, how well the computer will be able to use it, and how everyone else will be betting (as the odds are determined by other’s bets).

  3. Viral Frog says:

    Props to them for finding a way exploit an addiction to line their pocketbooks. -_-

    • Viral Frog says:

      Wow. Ignore me. Apparently I failed reading comprehension in the 1st grade. :/