Math Is Hard

[Gustav 6,406]

So I posted a status update a couple weeks ago in which I talked about a big project I'm doing...it's not done yet, so I think I'll take the opportunity to squeeze a few TPE out of a progress report (because how hard can that be? At the end of this sentence, I'll already be 14% of the way to the required word count, and I haven't even told you anything). 

 

I won't get into the why and the how, but I will get into telling you that this involves projections of every VHL roster. I'm trying to fully automate it, but I don't think I know enough to do that and I've settled for semi-automating it (and maybe at some point I'll be able to plug something in to take care of the rest. I like to think it's possible but I don't know). We're talking roster space, cap, TPE projections for every player, taking rookie and prime contracts into account and making sure we adjust for depreciation where necessary. It's taught me a couple things team-wise, too, because I'm not entirely up on that--Chicago is set to be in major trouble in a couple seasons' time, almost no matter what, and Helsinki has done a better job of resource management than I thought. But I digress. 

 

My point here is that math is hard. Even with numbers on my side for this kind of thing (yes, some of you have written articles that I'm going to cite), things still get thrown off. One would expect to see close to what I as an engineer would call a steady-state situation--players enter and leave at just about equal rates, and average roster and cap numbers across the league stay the same. But this is really difficult to capture accurately on a spreadsheet--just about every active prospect has a chance at making it up, but you can't just decide that someone who is earning now will stay earning in the future. So, you go in and assign everyone different "chances" of making it up and hope for the best.

 

The problem at this point is balance--you get roster numbers that look good enough, your cap numbers are also way off. You get cap right, your rosters get screwy. And, most importantly, something else I've realized is this: if you do "assume steady state" (I'll ping @JardyB10 here to give him a flashback or two) and make your projections just about equal the current numbers, then, a) congrats, you've successfully manipulated the data to give you what you want (and, to clarify, that is not a good thing), and b) what's even the point if you're trying to track the average numbers over time? If you knew they were going to look exactly the same when you finished, you could have just said that to begin with and saved yourself the trouble.

 

My second point is that I think it's basically impossible to track the future of the entire league. What I'm doing can be very valuable, and I think any individual GM running an individual team could find their own projections a helpful tool to plan their team's future. It's at the very least a decent guesstimate that I used myself as a GM and that always gave me a pretty darn accurate picture of what would happen to my team. I saw a (no disrespect) pretty bad projection of league sizes a while back and would like to caution against oversimplifying that sort of thing, even if I'm arguably overcomplicating it.

 

But lastly, I'd like to advise people to read carefully because there will be points where it doesn't appear that what I'm saying matches the numbers. There's so much scientific stuff out there that carries multiple conclusions. Sometimes, those conclusions aren't always obvious, and I feel like it's possible to wildly misinterpret the projections as accurate. 

 

It will be a good read, I promise--just with some caveats. See you when I get around to it.

 

Spoiler

 

[Gaikoku-hito 2,081]

1 hour ago, Gustav said:

Sometimes, those conclusions aren't always obvious, and I feel like it's possible to wildly misinterpret the projections as accurate.

Best comment when it comes to make projections as projections are simply the projections of the known but rather miss or cannot calculate  the unknown. Reminds me of a say,"We know what we don`t know but don`t know what we don`t know."

 

Sounds like an interesting project and maybe a bit complicated. I will be interested in reading since I have run similar projection projects on a small scale and in the end I came to the conclusion of: 

1 hour ago, Gustav said:

it's basically impossible to track the future of the entire league

 

Too many variable at play that cannot be calculated correctly or projected so tends to scewy the data to more of average or a "assume steady state"

 

I also interested in the semi-automatic aspect of your data collection since I tend to still take the long route of manual entering data that takes a much longer time. If I could ever figure out semi-automation or full automation then I am sure I would become a mad scientist running everything thing that I think of thorough a projection models; so maybe don`t tell me how you do this as I like to have a life outside the dungeon running data!! LOL!! 

 

Also by manually running the data I cross a point where I understand that I am not going to get the projection that I want due to many different factors or that is simple too complicated to continue to run the data as the benefits, time, effort etc.... don`t equal the returns of sad project.

 

1 hour ago, Gustav said:

I think any individual GM running an individual team could find their own projections a helpful tool to plan their team's future

The problem with projecting your own teams future is the fact that it takes you way from the "now"and into the future meaning that you are unlikely to make a bad trade that helps the team now rather than dealing with the future and the assets later. Since I have run such projections so I know that these projections are really meant to be broken if not shattered when great opportunities a raise. This is where you need as a GM or AGM to ask yourself a few different things about running your team. 

 

A. Are your projections making you make safe choices holding your team back from possible contending?

B. What is better being good over a long period of time or being great for a single season or shorter period of time?

C. Are you taking into consideration the Managers behind the players?

D. Is it fun? Playing it safe all the time rather than risky

E. Remember it is a game

F. STHS will F you when every it feels like!!

 

Projecting is okay but staying in the moment is more fun. Making mistakes or having a cloudy future means you have more opportunities to successful or an utter failure!! LOL!!

"We are as only as good as we are today but not as good as we are in the future"

"If you want the rainbow than you need to put up with the rain; ever if you are the one creating the rain!!"

[Spartan 4,402]

[Triller 1,659]

Very interested to see where this all goes!  The league provides a good data set that would otherwise take quite awhile to gather so the projection possibilities are very enticing! 

[Berocka 2,263]

8 hours ago, Gustav said:

math is hard

If math is so hard why do they teach it to 6 year olds

[BOOM 8,745]

Just because you're some kind of mathematical genius! 

[JCarson 726]

Article review: I always like reading Gustav's analysis of various areas of the league.  Now they do end up being long drawn out discussions that take sometimes months to complete but there is a sense of research, a general hypothesis that is then tested, not just accepting the basic fundamental assumptions of what is believed.  Gustav questions the status quo, challenging the self fulfilling prophesy  to add depth to the league and the overall discussion.  Keep up the great work, hopefully you can get your spreadsheet far enough along that it becomes relevant.  I give this article a 10/10.