How easy or hard is it to re-sign scouts? I'm assuming they don't have a morale system and there isn't a place they can elevate into like a GM role...
It is a little bit less of a tougher time to sign scouts as opposed to players because they don't have that morale system linked into them. However, they do reject based off of cash and years, as well as if I recall correctly if you are sending them out too much in areas they aren't supposed to be in. If you have a guy that's supposed to be an AHL scout and he isn't a good AHL scout he wouldn't like that, so there is a little bit there. It's not as sophisticated as the players, but we do hope to one day get to that point.
Do some scouts develop and do some stagnate, or do they all operate along the same experience gathering system?
In terms of how long it takes for a player to develop, it depends on their quality in that region. The longer they stay within there, they will move up in rank. It could take anywhere from an entire year to get to a relatively okay spot, but it is dependent on the scout and how poor they start off in that region. At the same time, as you keep them within there, they will start to decline within the other regions because they are now focused on developing in that one region.
At the beginning of scouting, I noticed you may not know a player's true archetype until you scout them. How off can those initial guesses from your scouts be?
It can be off, but only from a reasonable perspective. It won't be like an enforcer will show up as a sniper, mainly because in the real world you're going to know if the guy is playing that type of game. You may have a sniper show up as a two-way forward or playmaker, and enforcer showing up as a grinder or power forward, so it's within the realm of realism there. It works the same for pro scouts as well.
What are the benefits of having two scouts instead of one in a region? What would happen if you heavily focused on the WHL, for instance?
You'll definitely get more accurate results and you'll get more looks on players. As a GM you'll get one collective report based on that player. You'll get more accurate information in that regard, especially if you add a C scout and an A scout. The C and the A scout together would be a decent combination in that you'd get double the information on that player and it will be quicker for you to get that information. You'll also have more far reach for more of the players because you could assign 50 players to one scout and 50 players to another. You could also assign the same player to the same scout. There are benefits to it. Obviously if you do that, though, you are potentially missing out on another region so there is a bit of a risk/reward there. You may want to frontload the WHL but now you are missing information on the OHL. We tried to balance it in a way that that is a meaningful choice and you can't get all the information on all 900 players in a draft class. It's really up to you how you want to strategize that. You really may want to go all in on the WHL because you noticed that Central Scouting's ranked a lot of the players there, but you may miss out on the gem in the OHL as a result.
During a draft, I noticed the number one overall pick, who was an elite medium center, had a stat line that looked line with a player that was going to end up in a beer league as opposed to the NHL. He had like 8 goals and like 14 assists. I know you are still tuning, but how does that stat generation work under the hood?
What you probably saw was a junior player playing in a men's league. You have to be very careful when you're looking at a state line in that regard. If a guy's playing in La Liga, he's playing as a 17-year-old player against 25-plus-year-old men, so you want to pay attention to the strength of competition. Whereas if you were playing in the CHL, you're going to put up more points but obviously you're playing with your peers. That's why the stats looked off in that regard. You're actually having a kid play with men, which eight goals as a 17-year-old kid is pretty good in that league. Everything is tuned based off of the league. You'd actually notice if you had that same player and you placed him in the WHL he would actually generate an insane amount of stats. It's up to you to parse that data accordantly and it's just another tool for you to look at.
In regard to how we generate stats, it's a simplified sim engine in that sense. It's not anywhere as in depth as our actual sim engine, but it does take the players within the draft class. It doesn't go as in depth as, 'oh, this guy took a shot from here, this guy took a shot from there.' It goes 'hey this guy's overall is this, his age is this, this is kind of what we expect him to do.'
Do those junior or international league stats affect a player's progression when you leave them in those leagues once they are drafted?
It does in a sense in which their minutes being played does have an effect on them. If you are sending them back to junior, it does have an effect. It's kind of always had an effect when you do send a player back to juniors, especially when they aren't ready. They also do help you uncover fog of war information on your player just to keep a tab on a player. If they didn't play that year, then you actually wouldn't uncover data.
Have you made any changes in terms of how player progression works for prospects?
We have made some tweaks. You will notice that players in the past would grow all out of whack in terms of you'd get a guy with 99 speed and then a very bad acceleration. What we've done is try to make it tune with what an actual player looks like. Snipers will get better at shooting a little bit more than a playmaker and not grow as much in passing.
Talk about the gems and busts. Is that automated based on generated draft classes or does it vary how many are in each draft or is it a set amount?
It does vary. We generate draft classes on the fly so one franchise isn't the same as the other. Those busts and gems are different per franchise. They aren't predetermined. There is a formula behind it in regard to how good the scout thinks they are. The better the scout, the higher the likelihood that you are going to uncover those types of players, but typically you won't uncover more than like three gems and then a handful of busts. That's if you left it on auto-scout. If you manually scouted you may get to five gems, so there is definitely some benefit to manually scouting some guys, but there may only be 10 in a specific draft class. It's random but within a range if that makes sense.
Are those gem and bust designations guaranteed, or can scouts whiff on those?
They are guaranteed, and the reason we did that was to not have it be completely from a user perspective if I see this guy is a gem, I would want him to be a gem as opposed to being like I don't know what's going on? However, to counter that we made them super rare, so you're not getting a substantial amount. We didn't want it to be a very crappy moment where this guy isn't a gem, I can't trust my scout at all. But a crappier scout will have a very low likelihood of finding a guy.
When you put a player on your draft board early on, do the scouts pay more attention to him moving forward or do you still have to do that manually?
The draft board is more or less a user-based thing. It's more for the user to keep track of players they want. The scouts will obviously go scout players more often if they are within the top 250 of the draft class, so you should get relatively decent information on those players, but there is no bias to players on your draft board. It's more for pure tracking purposes for the player. In years past, we had a watchlist of five. We wanted to get it to 50 so they could really keep track of the players they want to grab in the draft.
How does this new scouting system affect trading during the draft?
As a result of the draft board, what we wanted to do was create shortlists for all of the CPU teams kind of like a user or ream GM would do. Every team takes into account their weaknesses, their strengths, how good the player is that is currently in that pick. Each team has a shortlist that they would want to work off of. Sometimes that list is as small as three per round, sometimes it's a lot larger – depends on the team. But they no longer take the best pick available. They will actually look at their prospect depth. That's scaled based on where they are in the draft. If they are in the top half they are going to draft usually the best pick available, but in later rounds they'll draft more for depth. Now that they know who they are targeting and who they want to get, if they think a player that they want is going to be picked ahead of them, they will try to move up in the draft. And if the player that they wanted is gone and they have no one left on their shortlist, they will be more willing to trade back and go to the next set of guys essentially. So, they're a lot smarter in that way. They're also more draft oriented trades. They'll do two picks for one based on the trade value for the draft picks.
When CPU teams create their shortlist of targeted players per round, is that based on where the players are slated to go? Is that the parameter for their valuations are or do they reach, too?
They do reach in some cases. Some of the picks won't be in line with central scouting. Sometimes guys will fall like five spots and they may be like 'oh, I'm going to grab that guy because that guy fell.' It really depends on the drafts and the team that is picking in that slot.
You have all this behind-the-scenes logic that dictates what these teams do, but it never really surfaces for the player. Have you thought of creating a narrative engine that takes the information of why these teams are doing what they are doing and translating it into a draft pick hot take from an analyst so it feels like the player can understand what the teams are doing?
Obviously, that would be an amazing thing to do. I think for this year specifically, just the amount of depth that we are adding we obviously couldn't get to everything that we wanted to do in regard to that presentation element. That's not to say that isn't something we may want to do in the future, it would definitely be cool to have that Madden-esque type of feel where they would have Adam Schefter show up in the draft as you are picking players. It's definitely something we would like to get to.
Let's talk about the introduction of pro scouting and fog of war. Why was this something you wanted to add into the game?
We've had amateur scouting in the past, and I really wanted to blow that up and make that feel in depth. And as we were designing the feature, we were throwing some stuff around and thought wouldn't it be cool if we had pro scouts and actually have a pro scout that actually means something? By adding something like fog of war and really flipping the game mode on its head on that regard where you are seeing every bit of information on a player and now you're not longer getting that information, we really wanted to make it a unique experience so that not every franchise mode feels the same based on the team that you pick, the scouts that you have on your team, you're going to get different results on players that you've scouted. You may overpay for players in trades as a result because your information isn't exact. There are lots of little things that in the real world the GM is going to make a mistake. In our old game, you weren't able to make those same mistakes because you could see everything so we really wanted to incorporate that real-life GM feel into the game. Having pro scouting and fog of war allows us to do that.
The fog of war also applies to your own prospects – coming into the league you don't have all the true information on them. How long does that fog last on your prospects?
It depends on how much that you play them. So, let's say I drafted a player that I didn't really scout well as an amateur player. Once he comes over we convert their amateur report into a professional report. Once they do that, you can then play them in the preseason if you would like if they have a signed contract. In that timeframe, you may get to say a three-bar accuracy and not a four-bar one. You may need to play this guy a little more so that he gets his nine-game trial in the NHL so you can really make a good evaluation. We're trying to get that same sense and feeling as a real NHL team would have to make a decision on a player. This guy may be a 78-overall player with three-bar accuracy, but he may be an 80 when you're scouting is fully accurate or he may be a 75. I need to play him more to get that information before I keep him on my roster or send him back down to juniors.
Will you get that information if they are playing in the AHL or is it only if they are playing up with the big club?
You will get that information if they are playing in the AHL. You will get that information but if they get sent back down to the CHL then you can't bring them back up.
How does the fog of war affect CPU trade evaluations? Do you have proposals where they don't have full information on their players so you can fleece them? Do some teams reject the trade, do the research, and come back and re-propose it?
They won't do that in the sense that they won't counter-propose a trade but it does affect the CPU trades in a manner in which sometimes they will trade for a player.... it’s more user oriented than CPU oriented but it does affect the CPU in a roundabout way. Let's say I'm trying to trade for a player. When they evaluate a player off your team and you are taking a player off your team they kind of have that similar knowledge. There is some of that but it's not as in-depth as the user side. The CPU doesn't have an advantage, they have some potential error in there, but it's not as great as it is for the user.
How long does your clarity last on a player from another team before the fog of war starts to creep in again?
We decay reports every 30 to 60 days. We also do regenerate a report for player every time you play their team again. If that player is in the lineup and you played them, you will generate a temporary boost to maintain that report on that player and get a little more accurate information. But by the end of the year if you never scouted a player again, you'll actually lose all the information. They will decay relatively quickly if you don't have a scout in that region scouting those teams.
I noticed you can scout line combinations. How is that beneficial as opposed to the other ways of scouting?
Right now, viewing lines would always be accurate. It's more for the users who play games, if they want to take a look at who's on the opposing team's lineup going into a game, it gives them that extra bit of information whereas in years past you could go to view lines and just look at their lines; like anything else you had full information on them. It's just to add a little bit of extra depth for the user that goes in and plays a lot of games so they can get a leg up on that team by scouting their line combinations.
Were there elements you were trying to get into the scouting mode this year you weren't able to get to?
This is one of the few features where I think we got a lot of what we wanted to get in. In terms of is there anything that we wanted to get in not get in, I don't think so.
Did you do anything different with player morale this year?
There's always fine-tuning we do based off player feedback, but nothing major. There's obviously bugs that we fixed with certain things but nothing major with player morale. Continuing to make it more performance driven than it was in the past when it was team chemistry oriented. We just played around with that a little bit more to fine tune it.
Any changes to how the CPU GMs handle team composition? I noticed in the past few years they are really bad about freeing up roster spots for trade flexibility. They're always up against that 50-contract ceiling.
Yeah, that gets better as you get deeper into dynasty. The problem is earlier on they are not as flexible because we're using the real-world rosters and we're importing players. By the time we get to September or October, they're kind of at that threshold. In regard to roster composition, we've done a few tweaks on how they look at players here and there. Nothing major but again it's more fine tuning when they will try to trade a player or what type of player they will sign. Very minuscule things.
I run into something every year and I don't know if its sim engine based or contacts demand logic, but you'll have a guy who is maybe an 80 like most players in the league, he only scores 10 goals a year. He's not super productive. He's a second- or third-line guy, and every once and a while you'll get these types of players demanding contracts way line with a high-skill player who generates a lot more points. How is that player's contract demands determined?
Every guy that you try to sign it depends on how old they are, so you won't see a guy that's done growing get to that point. They are basically banking on themselves to be better at a certain point in their career. What they'll do is look ahead and say, 'I'm going to be a good player, so I would like to have this dollar amount.' It's not always guaranteed so that's why you have to make that decision – am I going to pay this guy or am I not going to pay this guy? But it does have some logic in that they are trying to see if they are going to project into that potential they have and they'll demand based off that. Now, there's a lot of things that we've seen that we want to adjust with that. I've heard the community talk about this in the past and we just haven't had the chance to get to it, about it being more stats driven so that player even though he may grow into that 88 overall, he shouldn't be asking for that contract. There's definitely some stuff there that we should still address, but that's kind of what the thinking behind it is.
I noticed if you have a deep club like the Red Wings were the last decade, where your prospects are staying in the AHL until they are 24 and 25 and coming into that next contract, they'll be demanding a new contract like they've been a top-six winger in the NHL for a couple years even though they haven't played in the league yet.
Yes, and that kind of falls back to what I had just said – they are projecting themselves out. Now, they may not become good, they may become that 88 overall or they may become an 80. It's really up to you to sort of make that decision as the end user. Obviously, we want to make it better so it's a little tune with the real world and those expectations they have from a contract perspective.
There are a lot of moving parts to this deep new scouting system. Do you have any times for people just getting started so they get the most out of it?
That's a good question. Without giving away too much, knowing how everything kind of works, I would say really utilize your pro scouts to scout divisions at a time as opposed to individual players in some cases unless you are specifically trying to target a player. Otherwise, you may lose information quite quickly if you're not on top of the ball. In terms of amateur scouting, I would say try to diversify as much as you possibly can within each region because the way our prospect generation now is, good players are coming out of different areas quite often. You may want that All Cent Skin scout or that EBL scout just in case. Try to cast a wider net than you would have in years past where you focused just on specific areas.
NHL 19 releases on September 14 for PlayStation 4 and Xbox One. To learn more about the game, read: