My two favorite Candidates streams has been the Dojo and Matthew Sadler on Lichess. Seeing analysis from strong players without the engine is both instructive and entertaining. Viewers can still peek at the engine evals on their own if they want.
I'm a 1200ish player who is not even playing that much right now so kind of the equivalent of a sports spectator in a way and I'm fully bought in to the no-eval bar. Not from an improvement perspective but from an appreciation of the game perspective and that creates a very difficult question for content that is aimed at a population that has almost zero chess understanding because I truly don't know if there is any way to get that population to experience the beauty of this level of play, wether it's with an engine or not. I think people unfortunately fool themselves into thinking that the eval bar is a democratic/inclusive thing that gives more people access to the game in some way but I suspect there is a phenomenological issue there where some level of understanding is a prerequisite to experience that beauty. So that's definitely not an easy problem to solve for the wider population broadcasts that want to make more people tune in to chess events but at the end of the day when you're talking about making a narrative interesting you're talking about tension. If wider population broadcasts struggle to convey to their audience the tension behind moves I don't think the eval bar is going to help much there. I think you have to tell a better story about the moves that relates to things that they can actually understand like the player's life, current standing, when such a move or strategy was employed in the past, the riskiness of it, etc.
I think there should be 2 streams: 1. For those who don't care about chess(can be televised)- Eval bar visible to viewers, not visible for commentators 2. For those who care about chess(not televised)- Eval Bar not visible for anyone. Neither commentators nor audience. Commentators having access to bar/engine makes absolutely no sense in any circumstance
No offense to those who've been in the commentary biz, but I found Caruana's commentary in the recent WCC to be incredibly refreshing. There was an eval bar but Fabi was pretty much zoned into the chess while the other two were closer to professional eval bar readers (probably because chesscom valued Fabi's analysis more)
I don't even think that watching commentators without engine analysis, helps improve more. I mean if it's a tournament with multiple games at once, the commentators won't even have the time to understand the positions fully, especially in dynamic positions, where lines simply have to be calcualted, how is it gonna help me improve, if the commentators are as clueless as me, or only a little less clueless. I mean, I like watching games myself without eval bar, as long as I only focus on 1 specific game, but when switching back and forth between multiple games, the whole thing becomes shallow again. I always loved Peter Leko as a commentator, I feel like I learn a lot just from listening to him. And he usually analyzes the games without the engine (even though he sees the eval bar), while the co-commentator, also sees the engine lines. I like this format for commentating the most. One SuperGM or GM that watches the game purely with human eyes, while there is a second commentator who sees the engine lines as a blunder check, so to say. For the most part I also don't have the time to watch classical games in full length anyway, which is probably true for most people. I enjoy analyzing GM games without engines for my chess improvement, but for live events it's unfortunately not very practical. It's a fun format for sure but there are some preconditions that need to be given to make it work. That's why it don't think it will ever be mainstream. But of course that doesn't mean, that there is no place for it. If I have the time for it, I definitely enjoy it.
Good comments tobias. I would just offer one idea: if you and the commentators are somewhat or totally clueless about a position, this is reflective of what chess often feels like when playing. And then seeing how another (often strong) player approaches that position could be instructive.
@@chesscomdpruess In the example with Peter Leko I mentioned, it's pretty much the same, isn't it? He is looking for plans and does human analysis, explaining ideas and variations without knowing any engine lines, while having someone at his side that blunder checks with the engine. I agree, that if the commentators do pure engine analysis, it gets pretty boring for me. I mean, I could do this myself. I also agree, that if I have time to watch a game in full, I enjoy thinking about the positions myself and don't need an engine. It's definitely more fun. But when I'm at work or something and only step by for a few minutes every now and then to see how it goes, I definitely appreciate the engine analysis. For improvement purposes, doing it all on your own is probably still the best, at least once you are close to 2000 FIDE, you will only improve by thinking deeply yourself, cause a lot comes down to concrete calculation and proper visualisation. So I rellay don't think, it's better or worse for improvement purposes, but it can be more fun and exctiting, if you have the time to dive deep into it. But even this comes also down to the specific viewer and his playing level, so it might be different for different people. In the end there are pros and cons to both.
This debate could be used as an educational tool for all youngsters with political aspirations, because as far as I have seen in political debates not only in my country (Sweden) and abroad, the politicians have lost all their skills on how to debate. Exceleent debate, where the purpose was to solve a problem and explain ones arguments, not to any cost proving the opponent wrong. A respectful debate is essential for our civilasations survival. Bravo!
Great discussion! I think the Eval Bar is great for that snapshot to answer the question through a sports lense "Who is winning objectively (assuming computer is basically objective)/what's the score" type questions which is important to the casual viewer. But there is a level of surprise and suspense when you analyze without it, trying to dig into a deeper understanding of a move that may seem innocent or unimportant that later comes into play. Ultimately I don't think there is a right and wrong way to view it because broadcasts that approach it one way or the other are ultimately appealing to different audiences, have different motivations and ultimately get different benefits or detractions from it.
I think there can be different broadcasts that fulfill needs for different audiences. I turn on a bunch of different streams while the games are going on. One thing I will say is that chess engines, the eval bar, and online chess is the reason people are able to get into chess today. Anyone can start understanding the game by looking at lines and moves recommended by engines, and those moves even at low depth can tell you the ideas in the position. Great discussion.
I love the eval bar for entertainment purposes, but don’t take it too seriously. If two supernatural beings (God and Satan, say) with perfect knowledge were playing chess, then the eval bar would have meaning. But when two humans are playing, the eval bar can be very misleading. The main weakness of the eval bar is that it doesn’t tell us how HARD it is for human players to find the next best 50 moves in a row. It simply tells us White is, say, up 0.25, with best play on both sides. But if Black needs to find, say, 20 virtually impossible moves in a row to hold a game, then the eval bar while being correct in a cosmic sense is utterly useless in a real world human v human sense. An analogy: Imagine if the soldiers landing on Omaha Beach in WW2 were told, “With best play on your part if you zig and zag your way up the sand and make precisely the right moves for the next 50-100 steps, then you won’t get shot, won’t step on a landmine, and will reach safety. Eval bar says 0.0.’’
As a viewer who knows a little but not much (I'm 1500-rated), I sort of switch between streams. Those eval bar swings, even something small like +0.6, is something that keeps me invested if I don't wanna watch all 5 hours of a stream. However, I also do really appreciate the dojo style streams because y'all have more incentive to talk about positions as surprising, and really in a the way the players would be feeling them, even when the engine is saying 0.00. I'm with Kostya on this, the main broadcast should have engines. If I know a blunder has been made, I want the commentators to see the eval swinging, be surprised and then tru to figure stuff out. That's really fun. They shouldn't have lines though
Regardless of whether Hammer can effectively explain the engine evaluation within the framework of human understanding, as an amateur viewer, I find it difficult to shake off the influence of the engine once it's made visible to me. Hammer seemed to assume that the engine would enhance viewer comprehension in this conversation, but it has the opposite effect on me. Unfortunately, for chess events not covered by the dojo, it's rare to come across broadcasts without engine analysis.
Personally, as a chess player, I prefer to not have the eval bar there. But I get the argument in favour of it. Not sure a lot of people would watch a basketball game if the score wasn't displayed.
I much prefer commentary without the eval bar. I feel it kind of always puts players in a slightly negative light, so instead of praising their ideas, creativity and resilience we just compare how closely they match an unreachable standard of play. Saying it misrepresents chess is on point.
Great discussion. A related topic regarding how to make chess better for a TV audience: Time. Baseball is working hard to speed up the game. All sports are concerned with broadcast length. Audiences aren’t likely to watch a 5 hour game of any kind. What is the best Time Control to balance TV viewers and quality of play for big events that matter (prize fund, title, trophy)???
The thing I can't stand with Chesscom's recent "innovations" is that they show selected messages from livechat at the bottom of the screen. We have two or three highly skilled Masters commentating, who have been hired specifically for their experience in the game and their skill at explaining the position, but Chesscom insists on adding to this by some hiring a producer to make random patzers from chat appear on screen, usually with vapid comments like "This is a tough position, but I think Fabi can hold it". Literally every person watching the broadcast is watching on RU-vid/Twitch and can choose to show the livechat section if they want, but it's not enough for Danny Rensch's obsession with appealing to "the kids" and trying to make everything cool and interactive.
I don't think that there is a discussion on having an eval bar, if we ever want to promote the game that is absolutely a must. I think that the real discussion should not be if the bar should be there, but how it should be implemented. There are a lot of downsides to having the bar, so I think that its up to the comentators to explain it, but not having it is absurd
You can't be surprised by brilliance with the computer. One of the best moments of the dojo stream was everyone going crazy when Hikaru sacked the knight. With the computer you can only be surprised by bad moves which makes the game feel duller
I think David's main point about how chess with an eval bar misrepresents what real chess is is all true, But can a chess broadcast convey what the authentic chess experience is like to someone who doesn't already know what that is, just because they don't use one ? I don't think so. Because the only thing that lets people understand all the things david mentioned: the difficulty, uncertainty, the human aspect , etc is playing the actual game itself and even that will only get you half-way until you try serious classical tournament chess. Without that frame of reference chess commentary can do little to nothing to substitute for this even if it's excellently done with or without an eval bar. So why not let people enjoy chess as entertainment with the evalbar ? even if it's only about the suspense of "let's wait and see if they find the move the engine says". If it's enough to jumpstart someones interest in chess perhaps next year they are watching ChessDojo instead.
Hammer is on point. I won't watch it when they put commentators who just have a pretty face but clearly don't know what they are talking about. Having a panel beyond two commentators is annoying. Using engine analysis in commentating is pointless, as most people watching also have engines. I am interested in ideas from strong players, not engine analysis, etc.
I'm around a 1700 USCF and can see value in both. I have time to watch a classical game over 3 hours, more often than I care to admit, but people who aren't chess professionals largely don't. I can see David's point of the drama being heightened because of the uncertainty, but the amount of energy and attention required to engage on that level is a lot even for the commentators. Sometimes, I'm down to sit and stare at the position trying to figure out what's going on, but even with the help of commentators, I'm not going to really understand anything about top level games without the help of a engine if I'm not 100% focused. Sure I can think through the move I would make, and sometimes it aligns with the player, but the endless possibilities of chess can make it inaccesible, where if I see 3 or 4 suggested moves and continuations, I can focus my mental energy without feeling like I'm spinning in circles. On the other hand, if I'm analyzing a game between players rated under lets say 2200, I'd be less likely to want to use the engine, because I can reasonably expect to understand the logic of their moves better.
I also think the panel as a hole is overestimating the sincerity of people commenting things like oh Magnus or Fabi is an idiot for missing this move. I always assume those are just people trolling their fans, or just trying to get the goat of fans of a player that they are rooting against.
This could be easily done by twitch, if viewers have the option to close down the eval bar. That way improvers would have the option the close it down.
I want to point out that the Poker hand % winning chances isn't really equivalent to the eval bar. While it does map to "winning chance," unlike the eval bar, the poker %s do not in any way reflect what the players do, it's what the player are reacting to. The Eval bar maps directly to player actions in chess - every move can result in a change in the eval, but in Poker, the %s are the %s. Instead, in Poker, the thing being evaluated in terms of player actions is in the betting behavior, and there's no eval for that.
But i do love your streams. Prefer the dojo over any else. But i think the eval bar is unique to chess. It does help to get normies into the game. I think the thing that gets people to get lost watching is the multiple games at once or switching betwen them. My analogy would be if your trying to get someone into reading you wouldnt start with one book then switch to another when the first book gets slow.
I am generally in the no eval bar and no engine camp, both in annotating and in everything else. Sometimes envy the people who played chess in the era before the computer. But the argument from hammer that a commentator doing a top event with super GM's needs the eval bar to guide his commentary, without he has to use time and energy on calculating himself, made good sense. So the commentator can focus on conveying the event and the game in an entertaining way. And on another note. I absolutely love Hammer, hehe. Great sense of humor and the guy has some interesting things to say many times..
Watch out though. They might invent an AI that is trained on human analysis, psychology, every game each player has played, and all games in the database with what the player said and wrote about how they were feeling and there analysis after the game. Oh and it would also be integrated with a chess engine, and would know how to statistically predict the likelihood a human or a specific human would figure out the current position and what movies they might be more inclined to make given there style. So it will be able to exactly explain why Hikaru or Nepo or who ever made a mistake or a great move or what ever. Then the AI could go meta and also predict what they think popular human commentators would have said before the AI took their job. And how far off they would have been in the position. Then you can make the AI video live stream as a highly realistic computer generated hot chick showing lots of Cleavage. Maybe have them speak in an Eastern European accent to make it more authentic. And then unfortunately I think chess commentary would then be one of the next jobs to get taken out by AI. Lol.
Hammer and Kostya great as always. David argued like a snob. Everybody can enjoy the game as he*she wishes! If somebody wants to use it like candy crush, so be it.