Brad Wardell's Blog

Simulating the 2020 Election in Political Machine 2020

Published on Tuesday, February 18, 2020 By Brad Wardell In TPM Journals

One of the options in the game is to simulate the election cycle.  The player just chooses two candidates and sees how they do.  Now, you won’t get the exact result every time, but it’s fun to see how a given candidate will do if run by the AI. This mode helps us test against any unintentional biases in the issues.

Since no one knows how the election will go, it can be difficult to draw the line between when candidate A is simply more viable than candidate B.  In the 2016 version, we had to issue a patch. You see, Trump kept winning and that was impossible.  Goes to show what we know.

So let’s take a look a Sanders vs. Trump using the current data set we have (and we will be continually updating the data set between now and release).

Run #1: Sanders vs. Trump (AI vs. AI)

image vs. image

The stats we give to each candidate are subjective.  But since our goal is to make the game as accurate as possible, we do our best to be fair to each.


Early campaign


Mid Campaign


Late Game

From here we can see what mistakes the AI made.  For one thing, Bernie not even visiting New Hampshire is a glaring error.    Most of the time, the errors the AI makes are ones of taking certain states for granted.  Also, in this game, the AI did not focus on Medicare for All.


Run #2: Sanders vs. Trump (AI vs. AI)

Fundamentally, the AI players will make different choices each game.  They are the choices they think, at that very moment, will help them win but because each game is a little different (town halls and such come up in different places and we have a margin of error on the various issue data that we use to adjust the issues from game to game slightly.

Let’s see what happens game 2.


Mid-campaign in game 2.  One thing that the game concludes is that “Medicare for all” is a losing issue in Florida.  When he picks that issue, Florida goes red. 

We dug into this and (and again, this is a game) but seniors in Florida tend to have higher than average incomes and don’t tend to favor Medicare for all.  We will have to get more information on this.


Late game


Final map.

Anyone into politics knows that these results are impossible.  There’s no scenario where Trump is going to get Massachusetts or New Hamphire.  So what’s the deal?  It is a simulation issue or an AI issue or both?

In both cases, it’s an issue with AI.  Bernie just isn’t visiting these states.  Same with Wisconsin.  So then the question is, why isn’t he visiting those states?

Run #3: Sanders vs. Trump (HUMAN vs. AI)

So this time, I’m playing as Sanders. 

I also pick Amy Klobuchar as my running mate.



Mid Campaign

Playing as Sanders, I’m beating Trump pretty effectively.  However, I’m also not running my campaign as Sanders probably would.  I focus on Wealth Inequality, the Environment and Diversity.  I stay away from talking about medical insurance or college loans or guns.


End of game: Sanders victory

So what did I do differently than the AI?

For one thing, I didn’t ignore my core states.  I took advantage of AI Trump ignoring Oklahoma and edged Trump out there.  But the big thing was focusing on the wealth gap issue and addressing climate change and not so much on Medicare for all.

In the real world, Sanders won’t get Oklahoma.  But he didn’t actually need that to win.  He didn’t even need Florida.   I did manage to sneak Tennessee and Kentucky away from AI trump because, again, he didn’t start to visit there until it was way too late.


What we are learning

Trump is a difficult candidate to beat.  This is mostly because he puts certain states in play that most Republicans would not.  While I won Pennsylvania, Wisconsin and Michigan, I had to put quite a bit of effort to do so.  I had to give up Florida to ensure victory in those states.

From a game point of view, it all means there’s lot of replay value.  All the game’s key stats are in text files so players can play around with the numbers all they want and share them with others.

In the meantime, we continue to get more and more data and the team continues to improve the AI.

Political Machine 2020 #1: Being Pete Buttigieg

Published on Monday, February 17, 2020 By Brad Wardell In TPM Journals


The Political Machine is a strategy/simulation game.  What has made it so popular over the years is that while it is a game, it has a pretty serious and robust simulation running behind it.  Most of the data is collected from a wide array census results, surveys and demographic research. 

Every four years when we update the data, we are constantly surprised on the changes we see in different states.  The opioid epidemic is one example.  This issue wasn’t on our radar four years ago. Now it’s actually a pretty big issue in numerous states.

But in other areas, we can't rely on data. We have to make subjective judgments. Take candidate design for instance. Mayor Pete has gone through a few designs this time. Some people are easier to create caricatures than others.  He’s hard. At least that’s what the artists tell me. Mr. Buttigieg is doing well in the early primaries but should he become the nominee, can he win? 

Some candidates are harder to win with than others.   I would say Pete Buttigieg is a bit harder to win with than some of the others.  Here’s why: He has low minority appeal and for Democrats, there are a few states that become pretty hard to win as a Democrat if your minority appeal isn’t strong.  The problem however is that minority appeal is a subjective rating we give the candidates.  That’s why it’s so tricky.  What makes us qualified to say his minority appeal is weaker than say Joe Biden's?  We have some polling data but that’s really a cop-out on something that is simply a rating on a scale from 1 to 10.


So which states would be harder for Pete Buttigieg to win?


Michigan for one.



Florida for another.



Which isn’t to say he can’t win them.  It's just harder.



Overall, Mayor Pete is a bit tougher to play as than say Hillary Clinton or Barrack Obama but a skilled player can definitely win.

GalCiv III v4.0 data stuff

Published on Monday, January 20, 2020 By Brad Wardell In Galactic Civilizations III

The "big cleanup" is the plan for v4.0 of GalCiv III.  It's in progress but it's slow going because of developer availability (btw, we are hiring new C++ devs here..).  

So as I've discussed in the past, I'd really like to find a way to make modding more robust though that's easier said than done because of the dumb way we did our data structuring early on.

Basically you have lots of directories that the game pulls from which all depends on which expansions you have installed.  It's a pain.

When someone buys GalCiv III today from Steam, they actually get GalCiv III + Crusade + Retribution.  That's because we really want to focus on the assumption that everyone has Retribution going forward (or at least Crusade) which simplifies modding.  

I don't have an exact ETA on the v4.0 version other than it'll probably be this Spring.


Microsoft is the weakest link for the Surface Pro X

Published on Wednesday, January 15, 2020 By Brad Wardell In Personal Computing

As I mentioned yesterday, there is a really good use-case for the Surface Pro X as a device that uses Remote Desktop to your "main machine" when doing heavy lifting provided that it is sufficiently powerful enough to do light tasks on its own relative to say a Surface Pro 7.

Now, this is for my purposes.  YMMV.   

As you can see, for me, the Surface Pro X hardware has the edge over the Surface Pro 7.   

It's on the software support -- from Microsoft -- where things fall apart.  A lot of the benefits of the Surface Pro X hardware disappear when running Windows software under emulation.  This is particularly true on the battery life which just gets destroyed when its doing emulation. 

My serious work does include Adobe Creative Cloud (mainly Photoshop, Premiere, After Effects) plus Visual Studio 2019, Unity, Unreal, Cider, Nitrous, and lots of other development work.  But I'm not going to monkey with that sort of thing on an airplane or at a cafe or anywhere remote on any laptop (and if I did, I would use a Razer 15).

I am, frankly, baffled that Microsoft would even release the Surface Pro X without support from its own software.  For instance, MS Teams shouldn't be that hard to port and while it runs adequately in emulation, it's not the performance that is the issue, it's the battery cost of emulation. And this is an app that's running all the time.

So for now, I'll be sticking with my Surface Pro 7 as my mobile device of choice despite the fact that would really like the LTE support and battery life of the SPX.


A case for the Surface Pro X

Published on Tuesday, January 14, 2020 By Brad Wardell In Personal Computing

This is my work station area at Stardock when I'm working in the Object Desktop area of the company which is located on the third floor of the building.

As you can see, I have 3 monitors connected to a Surface Pro.  The reason I'm up here is so that I can collaborate in person with the rest of my team.  And the reason for the Surface Pro is that I frequently use it as a tablet for sketching on screenshots and designs.

The only downside is that the Surface Pro 7, while quite powerful, isn't really intended as a full-on development box.  It is more than powerful enough to run Office and tweak things in Adobe Photoshop and I've even edited videos in Premiere with it.  Even After Effects runs pretty decently.  But it's still no where near as fast as my main machine at Stardock which is physically located on the first floor.

At some point, Microsoft's Remote Desktop got really good at handling a multi-monitor setup.  It's had this capability for years but for some reason, I always remembered it as being a pretty crappy experience.  Recently, however, I decided to give it a shot and when combined with our internal fast network and the Surface Pro's high performance along with the crazy fast performance of my main machine I ended up with a solution that runs faster than if I were running the Surface Pro directly.

As a result, when I'm doing serious work at Stardock, I simply connect with Remote Desktop.  I haven't tried doing this from over a VPN yet (so no Coffee Shop test).  But even if it didn't make the cut when running via the Internet, the SP7 is more than capable enough to do all the work I could possibly need when using it cut-off from our network.

Which brings me to the Surface Pro 7.  While some of my friends in the tech community love their Surface Pro X's, I find that there just aren't enough ARM64 native apps yet to justify using it.  To be sure, the Surface Pro X is still more than enough for doing a lot of productivity work -- Jira, Visual Studio Code, MS Teams, Office and such.  But I wouldn't dream of doing any heavy lifting with it.

That's where the Remote Desktop pairing might really let the SPX shine.  First, there is a native Arm64 version of Remote Desktop.  Second, the Surface Pro X is substantially faster in video performance than the Surface Pro 7 which means it may drive my 3 high resolution monitors better (I do wish they had Thunderbolt 3 support however).  

As a result, whether I'm at home or work (my home gaming PC is a monster) the SPX would be able to use Remote Desktop to let me get real work done while still being powerful enough to get work done even when not connected to the network (which is why a Chromebook doesn't make the cut).

I won't claim that this use-case isn't a niche use.  But it takes the Surface Pro X from being, for me, completely useless to having a pretty viable use-case.

StarCraft Smurfs

Published on Monday, January 6, 2020 By Brad Wardell In PC Gaming

So I still play quite a bit of StarCraft online.  I am not great at it (low Diamond league) but I am good enough at it to know when I'm playing a Smurf.

A Smurf is someone who is deliberately playing in a league far lower than what they should be in and they really ruin the game experience.  

Smurfing isn't specific to StarCraft and there really isn't much that can be done about it that I can think of.  But it is one of those things that helps remind me as to why I don't want to develop games that rely heavily on competitive multiplayer.

If I play 10 games of StarCraft the breakdown is something like this:

2 games will involve Smurf accounts outright

1 game will probably be a smurf but can't tell because they immediately quit

1 game will involve some sort of obnoxious cheese strategy (as you go up in rank, this becomes less common ironically).

1 game will involve griefing. This isn't "cheese" per se because they aren't really trying to win as much as make the game unpleasant for you.

1 game will involve a ton of BM that forces me to mute the player.

4 games will be actual, legitimate games.

The only thing that saves StarCraft is that even though only 4 of the 10 games are real games, those games represent the vast majority of actual play time.

But the % of games that involve some sort of social-path spectrum player is sufficient to make me wary of ever wanting to make a game for that genre. 


The Star Control benchmark!

Published on Friday, December 13, 2019 By Brad Wardell In Star Control Benchmark

When developing the new Star Control game, we had an opportunity to see how far along the Vulkan and DirectX 12 tech was.  The answer, at the time, was not as far as we would have liked.  

In the end, Star Control: Origins shipped using only DirectX 11 because there just wasn't enough of a difference between DX11 and the new APIs even though the game sure could have benefited, especially on the planet, from the features of DX12 and Vulkan.

You might be wondering, what is the big deal about DX12 and Vulkan versus the old DirectX11?

There are a lot of differences but the big one is that DirectX 12 and Vulkan can have multiple threads sending commands to the GPU at the same time.  This means that complex scenes can be composited in various threads asynchronously so that you get the most out of your video card.

Now that most of us have several cores on our CPU, there's no reason to have only a single core talking to your graphics card.

The goal of this benchmark is to help users and our friends at AMD and NVIDIA and Intel to see how well their GPUs are doing with DirectX 12 and Vulkan.  Over time, we hope to see the DX12 and Vulkan scores pull away from the DX11 scores (and we'll be updating the benchmark as new DX12 and Vulkan improvements come in).

If you have any questions, don't hesitate to ask.


CursorFX cursors

Published on Monday, December 9, 2019 By Brad Wardell In WinCustomize News

With screen resolutions getting higher and higher there are many people who would like to have a bit more control over their mouse cursor.

If you don't already have CursorFX, try it out.  Stardock is currently working on a major new version of it but it will be compatible with the existing cursors here.

Here are a few of the most popular cursors here on WinCustomize:


Visit to check out the CursorFX gallery.


Post your rig setup

Published on Wednesday, November 27, 2019 By Brad Wardell In Personal Computing

I just got the Surface Pro 7 and I'm loving the power of USB 3.1 gen 2 (the USB-C port on the side).

I have everything connected via a single USB-C dock including two more monitors.

My main workstation is in the lower level.

Looking for WinCustomize volunteer news posters.

Published on Monday, November 18, 2019 By Brad Wardell In WinCustomize News

We are looking for some volunteers to help improve WinCustomize's news delivery.

Our mission is to bring you the latest, greatest news on everything related to the Windows user experience.  

Back when we were starting, what we now call "UX designers" were called "skinners" (not surprising that as what we do here got more respectable a more refined name would be used).

What are the kind of things we want to cover? Here are 3 examples.

  1. Interesting designs created around the net.  Example: this by this person.
  2. Interesting new Windows specific hardware like this review of the Surface Pro X.
  3. Interesting things created by the WinCustomize community you want to draw attention to.


  1. Must have had an account here for at least 90 days already.
  2. Must be a good writer.

If you're interested, DM me.


718 pages 1 2 3 4 5 6 7 8 9 10 Next