Open Looks

Just a heads up - Franchise mode is not in the demo. The demo is four teams and a game of hockey. This devlog is about stuff in the full game.

And it is work in progress. It is built and it runs, but it has barely been played by a person (basically just me).

With that said, this is one of the biggest things in the game that nobody has really seen.

Seven rooms

The mode has seven rooms and the shoulder buttons move between them from anywhere, so you are never backing in and out of menus over and over again.

The hub is where most of the core “gameplay” pieces live. Play the next game yourself, sim it, or sim to the playoffs. That last one exists because I do not want to play every game every season, especially testing out longer franchise runs.

The room

This is the part most likely to change.

Every player has a general disposition: leader, workhorse, etc. On top of that sits a mood that moves up and down. It comes from where you play him, whether the club is winning, whether he is producing, what he is paid against what he is worth, who is coaching him, and whether he is hurt.

All of these things affect the game directly. Mood reaches the ice as effort when you play a game, it moves the result if you just sim, and it reaches the money, because a man who is unhappy asks for more to re-sign or won’t give you the time of day. If you bury somebody at the bottom of your lineup for long enough or send him to the minors, he asks for a trade.

Every rival GM can see that he asked, and he fetches less because of it. That is the cost of letting a room go bad, and all of this is modeled after a certain current Detroit captain who needs to go away.

Beside it is line chemistry, which is two separate things on purpose. Fit is what two men are: a passer next to a shooter, a lefty next to a righty, etc. Fit sets the ceiling. Time is shifts actually played together, and it builds over time so keeping lines consistent is helpful for your team.

Your head coach you hire will build your lines for you. He is biased by his own system, so a defensive coach rates one way, and a run-and-gun coach rates another.

If you don't want to manage lines yourself, there is a COACH'S LINES button that sets them all automatically.

You also pick your own captain now. That used to be automatic, handed to whoever had the highest overall. Which as a Detroit fan, I know is a bad idea.

The trade room, which I keep changing and redoing

The honest version of this section is that I built a trade room, used it, hated it so much when in actual use. One screen was doing three jobs and it sucked at all of them. Offers came in as a slim row that popped a card on top of the builder, and everything was just being crammed into one screen as the trade feature grew.

It is three separate tabs now.

Those three numbers are there because they are the three things that make two offers easily comparable.

The block came from annoyance from my own testing: every man in the league was equally available, which is not realistic unless you are 2026 Detroit. So there are no-trade clauses now, and asking somebody to waive is a real conversation. He weighs where he is going against how he feels about where he is, and he answers the same way every time you ask, so you cannot keep trying until he says yes. A yes is a yes for that club and no other.

You can also negotiate a no-trade clause into a contract when you sign or extend somebody. A man who wants one takes less to get it and charges more when you refuse.

Now, the trade deadline. The weeks in front of it now give information out. Rumors land on the wire, more often as it gets close. They are read off the actual block and the clubs actually shopping, so a club "in on" your winger is a club genuinely short at that position. And a man in the last year of his deal is a rental, which means he is worth a lot to a team chasing a cup and nothing to a club that is not.

Scouting and the draft

Scouting is an in-season job, not a thing you do the week before the draft. The class is generated a year ahead and you watch it all season. You have four scouts, you send them at specific kids for a number of weeks, and a scout can only follow one player at a time.

What you see on the board is a range, not a rating. 42 to 54 overall, potential somewhere between A minus and A plus. Your scouts narrow the range on the kids you spend them on.

The draft itself runs pick by pick, and every club's turn is a screen you press through, but there is an option to quick sim to your next pick.

The front office

You hire and fire the whole staff, not just a head coach. Goalie coach, power play, penalty kill, minor league, head scout, fitness. An empty seat is not neutral, it is work nobody is doing. Each coach does affect your games, your development, everything, so they do matter and balancing your spend is important.

The finance side is deliberately light. There is a soft cap and there are contracts, but this isn't an accounting sim, because I want the fun and challenge of being a GM, but without needing a CPA to understand the spreadsheets. One thing to call out is the league cap and your owner's budget are two different numbers, and his is lower. The cap stops you outright. The budget just disappoints him, and sitting over it costs you money and standing.

The wire is where you can see what is happening across the league. Streaks, milestones, injuries and returns, etc.

What is not done

Year one records screens are mostly empty by design, which is correct but reads as unfinished until you have played a season. There are certainly more layout bugs because 24 screens is a lot of screens for me to have looked at and found every single issue.

One more thing

Everything above is you against 23 computer GMs. But what if there was also a version where the other GMs are actual people. A league that lives on a server, everybody runs their own club, trades are between real people, and when two of you are scheduled against each other you sit down and play the game.

That is being built. More on that one later.

If you want to argue with any of these decisions, do it!! The demo is free if you want to see how the hockey itself feels: everything above sits on top of that game.

More entries as I keep building. Wishlist on Steam.

The demo went out on Monday. As of this morning it has been downloaded 1,051 times and 153 people have actually launched it. A week ago the number I was quoting was twenty-two. I am not going to pretend a thousand is a big number in Steam terms, but it is a lot of people I have never met playing something I made at my kitchen table, and I have not really got my head round it yet.

Thank you. Genuinely. And a specific thank you to everyone who left feedback in the game or left the anonymous play data switch on, because almost everything below came out of one or the other.

What the numbers actually said

The honest headline is that the median person plays for eight minutes and does not come back. Eight minutes is almost exactly one game in the demo, since periods are 1:30. So people are turning up, playing a full match, and leaving. Only about one in three ever launches it a second time.

The other one that stung: of everyone who finished a match, only around one in seven pressed REMATCH. I had built a thing that offers you a spicier rematch with a house rule turned on, and I had gated it so it only appears once you have already played two games in a row. Almost nobody was reaching game two, so the feature meant to create the habit was only ever shown to people who already had it. It fires after the first game now.

Same story with 3-ON-3. It was sitting behind two menus, and out of everything people started all week it was picked six times. It is on the main menu now.

The freezes were real

Some of you were getting multi-second freezes, usually right at a faceoff. I could see them landing in the data but not why, and I had already shipped two fixes aimed at causes that turned out to be wrong. The uncomfortable part is that the second fix went out and the rate did not move, which I only know because I went back and checked instead of assuming.

So this build stops guessing and adds the instrument I should have had first: the game now records how much shader compilation happens during a stall. If the next one comes back with a big number, it is compilation. If it comes back zero, I can stop looking there entirely. Either answer is worth more than a third guess.

Online is worse than I thought

I finally measured it properly. Across every connection attempt in the first week, roughly nine in ten failed before ever reaching a game, and almost all of them failed the same way: the person joining could not reach the person hosting. That is a home-router and ISP problem more than a game problem, but it is still my problem to solve, and "some routers will not let you host" undersold it considerably.

I have not fixed it. I know exactly what it looks like now, which I did not a week ago, and it is the next thing I take on.

Everyone on the ice got redone

Skaters, goalies, the ref, the coach, the crowd, the zamboni. Jersey stripes are real textures now instead of projections, the goalies have proper pads and masks with actual designs, and both sticks were built from scratch instead of the placeholder assets I had been living with since the start. The models are simpler than they were too, so lower end machines should pick up a bit of performance.

It is not finished. It is closer to the thing in my head than it has ever been.

A few numbers that made me laugh

• You scored 153 goals and hit the iron 130 times. 85 crossbars, 45 posts.
• Hardest shot on record: 108 mph. Longest goal: 26.8 metres.
• 426 hits thrown, 47 of which ended with somebody folded into the boards.
• There were 23 own goals, and 18 of them were scored by a human being rather than the AI. I have looked at them and I have no notes.

That own-goal number is funny and it is also probably telling me something, since every single one came from a scramble in the crease. Adding it to the list.

One thing about the data

Anonymous play data is on by default from this build, with a notice on the first boot and one press to turn it off. No name, no Steam account, nothing that points at you. It was opt-in before and I was only seeing about a third of players, which is not enough to trust any of the numbers above. I would rather say that plainly here than have you find it in a settings menu later.

The full change list for this build is in the Discord. Everything else, as always, I want to hear about.

The Barnburners demo is live on Steam, free, as of this morning. Here it is.

Four clubs - Chicago, Seattle, Montreal and Toronto - playable in exhibition, on the couch or online. There is 3-on-3 as well, and a few drills if you want to learn the controls first. Periods run 1:30, so a full game takes a few minutes. It is built for “one more game”. Two house rules are unlocked: MEGA HITS and BIG HEAD.

Franchise mode is not in the demo. That one comes with the full game, which is heading to Early Access later this year.

What I actually want from this

The point of putting it out this early is to find out what works before the rest gets built on top of it. If something feels wrong, breaks, or just is not fun, I want to hear it - that is the whole reason the demo exists. The Discord is the fastest way to reach me and I read everything.

Known issues

• Hosting online. Some routers and ISPs will not let you host. The game tells you within about ten seconds if that is you - join someone else’s game instead. A proper fix is on the list.

Anything not on that list, please tell me about it.

On June 14, I got the steam approval to put a playtest build out online for the first time. Twenty-two people have opted in to tracking since, and 50+ people have installed according to Steam. That is not a big number, and I am not going to pretend it is, but it is more than were playing it in May, and that's pretty cool.

Since that first build there have been sixteen releases, from big adds to small patches. Here is what actually came out of it.

The things you asked for

My last devlog put the feedback up on a public page with a status pill on each theme, and I said the pill flips when I ship it. Update on some of those:

• "It's too fast to follow." The most common note by a mile, six people said it. There is a GAME SPEED setting now, and the default is SIGNIFICANTLY slower than what I originally had. I also added reaction time to the AI decision making. They no longer pass the puck around like crazy and instantly. They "take time" to think and also make mistakes.
• "The defense is relentless." Defenders play real zone coverage instead of all 5 players chasing the puck, and they run actual systems now - breakouts, forechecks, a power-play overload. I tried my best to implement some real hockey plays and I'll continue adding in more.
• "I can't keep the puck on quick passes." This one was open for a long time. A pass now gets caught in a cone in front of you instead of snapping to your stick from any angle, and a puck you win off a battle sweeps in rather than teleporting. Looks more natural to me. Also the overall speed of the game helps this a lot.
• "The skaters don't feel like they're on ice." Still in progress. I've changed how turning works while boosting and how stamina affects turn radius. Will love to see how the current changes affect this. I don't want the game to feel as floaty as the current EA games, so its a fine balance.
• "The tutorial needs clearer prompts." FIRST SHIFT got rebuilt, it pauses on each instruction, and the button prompts match the controller you are actually holding.

The things nobody asked for

Fights got torn out and rebuilt as an actual fighting game, with combo strings and a knockout finisher. There is a Training Camp with drills (only some of them make the demo). There is a replay theater and a photo mode, with them recently getting combined into one vs separate. There is couch multiplayer that actually works (when I first put it in, I didn't test and it would crash the game), two players against the CPU or four pads for 2v2. The game UI supports ten languages (thanks google translate). The graphics settings calibrate themselves off a real match on first boot instead of guessing, hopefully alleviating some of the framerate issues I can see on my end.

The whole UI got swapped out too. It started as a pixel, 8-bit kind of look, which I liked for about a month, but it made the game read as a retro throwback instead of what I feel the game is more leaning towards. It is a bit more early 00's in my opinion, also just easier to read. You can see it in the two shots at the top.

The skater models have gotten a huge repass as well, even if they still look bad. Their skates FINALLY plant on the ice instead of sliding through it, their bottom hand actually grips the stick instead of hovering near it (but still is a bit rough), and the stick blade grounds on the real geometry of the ice. I keep on redoing animations all over the place, trying my best to make things look a bit more natural. None of these are features but all of it is the difference between looking totally janky and unfinished, and only partially janky.

And there is a commentator now that will ship in the demo build, calling the game in a real voice. I'm excited for that to go out with the demo and see what people think. I personally play with it off, but that's because it's me, and its really weird to listen to myself.

What I got wrong

A boatload. Pucks stuck to an invisible shelf above the net when I swapped net models and didn't update the collision. A goalie could get announced for scoring a goal that was deflected in off him while he was the one getting scored on. The crossbar sound played when you hit the post. Two centers showed up for the same overtime faceoff. A sixth skater would wander on after a fight and never leave. Sometimes that'd happen multiple times and a team would have like 8 players on the ice.

Every one of those was found by somebody playing, usually somebody who just said "the puck got stuck" and left it at that. That is genuinely all I need. I have the telemetry and the replays to find the rest.

The demo lands Monday, August 3rd. Six weeks ago this was twenty-two people. If it goes badly this time it will go badly in front of considerably more, which is the whole point of doing it in the open.

I'll probably be ending the playtest after the demo is out and restarting it for a smaller group as I work on the rest of the features. The past 3-4 weeks have solely been on polishing gameplay and what is included in the demo, so the rest of the game really needs some love.

Barnburners is made by one person - me. So one thing I think is cool is actually putting my actual numbers on the internet, live, for anyone to see.

There is a new page on this site called The Numbers. It reads straight from the game every time you open it, and it shows the real state of the playtest from those who have opted in to sharing data. It shows how many people have installed it, how many matches have been played, how many goals and hits, and how many of the people who left a rating liked what they played. All anonymous, all opt-in, and no cherry-picking.

The part I care about most is the feedback. After a game, it'll sometimes prompt you to leave a thumbs up or down and a note, and those notes have shaped a bunch of updates. A few things came up over and over: it felt too fast to follow at first, people wanted to control the goalie, and a lot of you want the skaters to slide and glide more like they are actually on ice. The page groups those into themes and marks what I have done about each one: shipped, in progress, or still on the list. When I ship something, the pill flips.

Behind that public page there is a messier one only I look at. Mainly the same telemetry, but the whole picture. Which GPUs and resolutions people run, how the frame rate holds up build to build, which modes people bounce off of, and the crashes. Building blind sucks, so this is how I decide what to fix and change instead of guessing.

That is the whole idea. I would rather show you the real thing, issues and feedback and all. The demo is coming soon, you can request access to the current playtest (which is just the demo right now in its finishing polish stage), and the game is on Steam to wishlist, and if you play it, tell me I got it wrong. That is the point!