Huge Developers Update: Total Pro Baseball 2004

[ScottVib]

Shaun's latest update..

http://www.puresim.com/tpjournal.aspx

[mtaystl03]

Yes! A rule 5 draft! Im getting it.

[ScottVib]

For those who haven't clicked the link yet:
-----------------------------------------------------------
Thursday, January 8th 2004
Focus - Sometimes Less is more

Sorry its been a few weeks since I posted an update, but I have been deep into development. I have also finalized "the list" of features that I am implementing for 2004 and it is huge. As I mentioned in the last update I have been focusing almost exclusively on the various game engines under the hood.

Before I go into all that I'll say that my current kick has been "addition by subtraction". This has been quite a journey -- I NEVER expected to make a radical shift in philosophy with 2004 but what I have arrived at amounts to just that. While Total Pro 2004 will retain some of that PureSim flavor, it is really turning into an entirely new game. Basically, I sat back and took a look at a bunch of PureSim game features that while on the surface appeared to add depth to the game, really didn't -- they just became "tasks". Some examples: Spring training focus, Hiring a manager, Scouts, Team Chemistry. I have decided to remove all these features and refocus on a much more streamlined game experience. Instead of depth in those areas I am adding depth to the GM and Manager aspects of the game, with a keen focus on AI, stats and the Player Career trajectory development engine. OK, so lets talk about what I am doing / have done:

No more scouts

As I mentioned in the previous updates they are gone. This has resulted in a much tighter game experience and as I expected an AI that is making decisions based solely on statistical performance (past and present) of players is yielding a much tougher computer opponent. The catch here is the initial draft. How can the AI make a determination on player value when there is no previous performance to evaluate? I am still on the fence here, and can go a few different ways I'd love to hear from you on which approach you prefer:

Keep ratings like contact, power etc. but simplify the scale (e.g. 1-10 or A, B+, C etc... This way you'd still get a general idea of a player's talent for the initial draft.
Run each player through the simulation engine to build up a representation of their average Major League stats. So when doing the initial draft, you'll see a 28 year old LF with .287 27 HR 31 Doubles etc... These stats represent his avg season in the pros. For amateurs/young players these stats might be post-fixed with (AAA) or (AA) or (College) meaning they represent how that young player has performed at a certain level up to now. Basically the game shifts from being ratings-based to stats based. Of course some ratings stay around (for example Speed and Velocity) because those are truly accurately measurable.
Keep the ratings exactly like they are. Now without the scouts this means they are a very accurate representation of how good a player is. Too accurate you wonder? Well, that is the crux of the issue. Interestingly, with a dynamic stats and player development engine variances happen "organically". So, things are certainly dynamic enough that they are not formulatic. Just because a guy has a 90 POWER rating doesn't mean he is a lock to hit 40 HR. That said, it is likely he will be one of the top power hitters.
[/list=1]

I am totally on the fence with this decision, leaning towards either option 2 or 3. What do you think?

Career Trajectories

While PureSim was good here, I did some intensive analysis and research and have come up with an engine that has better feel. Sometimes in PureSim the talent swings were just too drastic. The new engine allows for even more variety as well. That said, variety should not mean "random"! Some guys will peak real early but never really get better, others improve very slowly and steadily yet maintain a high level of performance for a long period of time etc.. This is probably the area I am the most proud of because it is critical to the long-term stability of the statistical results. Balancing how the stats look after 50 years use to be a monumental challenge -- now things are totally stable, yet dynamic enough to add variety.

New player generation algorithms

I spent a whole weekend just doing statistical analysis of players and their attributes (both PureSim and the Major/Minor Leagues. I ended up building a new player generation model that has much higher fidelity and variety. Another cool innovation here is the game now gives you an option to generate fictional players that approximate the type of players from a specific time period. So, say you want to build a 1909 association, you'll have the option to generate players that match that era very closely (very few homers etc..) Another key advancement here is I feel like I have a dynamic enough model where I am seeing some "Rob Deer" type guys (e.g. hit .230 but crank out a ton of HRs). Not a lot, in fact just enough to keep it interesting and on par with the real world. In PureSim there tended to be a little too much correlation with high AVG in those cases.

The real depth in 2004 - GM / Contracts / Rosters

OK, above all else this is where I want to really focus the "fun factor" of the game. My goals are a bit lofty, but I am going for it. Here is what I am trying to get in there for Total Pro 2004:
Multiple Levels of Minors - You will have A, AA, AAA rosters. These rosters will not have to be micro managed. You will be able to make high-level decisions like playing time, but not have to keep them at 25-man size or do lineups. Stats are simulated using the core game engine, but in abstract fashion. The key part of the game here is that it is fun to develop players and this will capture that.
[LIST]
40 Man Roster - You must maintain a 40 man roster, these players also must be under contract and directly hit your payroll. You may not call up a player unless he is on the 40 man roster. Players not on the 40 man roster but in the minors are on what will be referred to as a "Minor League Contract". Of course you can only bury a guy in the minors for a certain number of years before he can leave (or get plucked in the Rule 5 draft).
September call ups
Waivers and Options. This will work just like it does in MLB
Rule 5 Draft
Compensatory Picks.
10/5 rule

So, as you can see these features open up whole new challenges to being a GM. In fact with a stats focus (instead of ratings) and all these GM features I think the game will really simulate the role of being a big league GM very well. What do you think? Is this too much? Am I crazy? :)

Other Stuff

This is already getting too long, so I'll close with a list of other things I want to get in:

Speed! Speed! Speed! - Probably the biggest complaint about PureSim was that it was too slow. Huge gains have been made here. Creating an association takes 1/3 the time and sim speed has improved to the point where I can sim a 16 team league in 8 minutes on my 1.9ghz Pentium. I track (and store) the results of every pitch, so Total Pro will never be the fastest horse in the race, but sim times are quite acceptable given all the stuff that is being managed. The act of jumping in to an association, creating it and playing is just much more streamlined now with all the "noise" removed.
Conversion from points system to $ (I know, I know I should have done this long ago!)
New User Interface - More real estate, right click menus, customizable etc...
Conversion from points system to $ (I know, I know I should have done this long ago!)
Better, more interactive off season Free Agency signing period
More features for folks that want to manage individual games.
More stats - In general I am just trying to include more stats wherever I can. Especially sabermetric stats.
More Awards - ROY, Minor League Player of the Year. Individual awards for each league.
GUI for tweaking some important engine settings
Multiplayer online league support


Should it be called Total Pro Baseball 2005? :)

Astute readers may be wondering what all this means for the release date of the game. While it is not my policy to give release dates I will say that this new direction will push it way back. Total Pro will likely not be available until some time in the middle of the summer. I guess it boils down to this, I could just tighten up the existing PureSim 2003, add a few new features and release it as Total Pro Baseball, or I can use the core PureSim engine and code and build a really fresh game (which is what I am doing). I just hope you can wait til the middle of the summer (July?) for all of this. What do you guys think?

Edit - Developers Update Thread at the Total Pro Baseball Forums (http://www.400softwarestudios.com/invboard/upload/index.php?showtopic=47573)

[Bonegavel]

Glad to hear about the scout death. Shaun must have read this thread a while back :D

My scout rant (http://dynamic2.gamespy.com/~fof/forums/showthread.php?s=&threadid=13344&highlight=scout)

[johnnyshaka]

Hmmm...sounds pretty cool...like the Rule V draft, the compensatory picks and the 40 man roster stuff.

Just might have to check it out...

[Ramzavail]

I absolutely love Option 2 - stat based rather than rating based

[JasonC23]

Wait a minute...he's REDUCING the feature list to focus on making the CORE product work as well as possible?

Are we sure this is really a .400 product? ;)

[Bonegavel]

ouch!

[SackAttack]

I'm all over the idea of stat-based rather than ratings based.

Do I WANT to wait 'til summer for this? No. Will I? You betcha.

[Ksyrup]

I think that's a pretty good idea, actually. If FOF had been released at the beginning of the football season and TPF came out, say, in November, I probably would have had time for both. But now, it's just FOF until baseball season. With TPB and OOTP, we'll get OOTP in March then this 3-5 months later, meaning I'll probably have time for both.

[Karim]

I've fired an e-mail off to Shaun. Stats based is really the way to go.

[MizzouRah]

I love option #2 as well.

Is it finally time ootp has some competition?



Todd

[HornedFrog Purple]

This looks great. I can sure wait until July, Shawn Sullivan's track record speaks for itself. If something doesn't get in he will tell you immediately and his reasoning for it.

I wait with anticipation. The best thing besides the certain features and design decisions is the increase in speed. Cool cool cool.

[MrIllini]

will you be able to port OOTP leagues into this game?

[Danny]

Two different games and companies, so Im sure you wont be able to.

[FBPro]

Originally posted by MrIllini
will you be able to port OOTP leagues into this game?

I would certainly doubt it.

[Easy Mac]

So are you going to have a paying beta for this one too (with a "free" full game when it is released) or just give it to use when it is ready, with everyone paying full price?

[Glengoyne]

Originally posted by Easy Mac
So are you going to have a paying beta for this one too (with a "free" full game when it is released) or just give it to use when it is ready, with everyone paying full price?

Shaun is NOT Arlie. Don't lump Puresim/TPBaseball in with TDCB and TPF.

That said, I am sure that TPF will be fine, just like TDCB is. Arlie may not have perfected release, but he does address problems that arise.

But then again I am in the "I don't give a rat's ass whether or not I can try for a two point conversion" camp.

[rjolley]

Glen: I think he was refering to the program Shaun had last year for PureSim 2003 where you could buy the beta at a discounted price and be a part of the beta program.

And to answer, I think Shaun said he would not go that route this year. You may want to post on the TPB board to be certain.

Shame, I liked the idea of that...get the game early at a discount and help shape the final product.

[Daimyo]

I tried the demo of PureSim last year and thought that it absolutely killed OOTP in terms of simming baseball, but felt way too bogged down with the other stuff to enjoy the game. Definately a good direction to go!

[Glengoyne]

I had completely forgotten about that bargain priced beta plan. It was a good idea.

I am just so used to people grasping at just about anything to bash the folks at .400 studios, I misinterpreted Easy Mac's comment.

[Barkeep49]

Shaun and .400 have indicated that there will be a traditional beta process and I agree it's too bad since I think the open betas made Puresim a stronger game.