Sunday, March 20, 2011

If you haven't played Osu!, then you don't know your mouse. This is what I realized, about 2 years after that fateful day when I discovered it while looking around the interwebs for MMORPGs.

Through googling and random experimentation I've learned quite a fair bit about mouse input in general. It's certainly not everything, but it's enough that I am afraid I would forget some of it. So here I have decided to put down a few points for my personal future reference, as well as for anybody who might wish to get to know mice better.

*I originally wanted to post the software part together, but it seems that the hardware part is quite lengthy by itself. Maybe I'll post the software half later, maybe I'll be too lazy to do it.

EatNow's guide to Mouse Input

What this is (and what this is not)
This is a short collection of pieces of knowledge I have gathered about optical mouse input. It is based on what I have learnt from the music game Osu!(basically a music game involving clicking on lots of circles, and the closest you can get to a mouse test), and it is of particular relevance to real time strategy games and any other . Since I do not play first person games in general, the things I say here may or may not apply.

This also has very little to say on wireless mice, which may be your preference. As far as I am aware wireless mice are subject to the problems of weight (due to extra batteries), latency and other connectivity issues. However I haven’t used one myself and therefore do not know anything beyond this.

********************

Hardware

Do a simple google on “optical mouse recommendation”, and you are bound to find lots of forum threads with people posting things along the lines of “it’s the skills not the mouse” and “any 5 dollar default mouse will do”, etc. Well to a certain extent this is true (see under DPI section on small screen resolutions) but after trying out Osu! with different mice I believe I can safely conclude that the mouse does matter, and quite a fair bit. The difference is extremely obvious on Osu!, and probably not as noticeable in more complicated games, but a better mouse definitely makes you a better player (or at least it is an important prerequisite to your improvement) in any top down game involving clicking with moving cursors. For mouse input hardware is what you will be spending money on, so it is probably much more important to look at the following sections first if you are looking to buy a mouse.


Friction

The reason why I put this section first is the tiny amount of attention it gets relative to all the other “specs” (such as DPI, refresh rate, etc. which will generally appear somewhere on the product packaging), when it is actually one of the most important limiting factors to input accuracy. Because of this (and the fact that you can’t feel the friction of a mouse through the packaging) it is a little difficult to make judgment in this aspect of a mouse. Now a little on the physics behind friction (you may wish to skip straight to the section on practical tips if not interested):

Firstly, there are two kinds of friction: static friction and kinetic friction. Static friction applies when the two surfaces in contact are not moving relative to each other, while kinetic friction applies when the two surfaces in contact are moving relative to each other. Static friction is higher than kinetic friction, which in the context of mice means that it would take more effort to get a mouse to start moving than to keep it moving. This gives a sticky feel to the mouse when you start moving it, as well as when you try to reverse its direction or do sharp turns (there is a brief moment during the reversing of direction where your mouse is stationary). This is bad for your mouse accuracy, so you would want the static coefficient of friction to be as close to the kinetic coefficient of friction as possible. This is quite a bit more important than having an extremely low kinetic coefficient of friction (which combined with a high static coefficient of friction, will just result in a lack of control with the cursor overshooting the target most of the times).

Secondly, the coefficient of friction between two contacting surfaces depends on the material and texture of the two surfaces in contact (pretty commonsensical and obvious to you I hope).

Thirdly, the actual force of friction (or more accurately the magnitude of the force of friction) is given by the coefficient (either kinetic or static) of friction multiplied by the normal reaction, which is basically the weight of the mouse plus the amount of downwards force your hand exerts on the mouse (which may be negative if you are lifting the mouse, or very positive if you press down hard on the mouse).

What the above 3 points translates to in actual scenarios:

Point one: If you happen to have access to sample hardware for testing, what you want to look out for is both the static and kinetic coefficients of friction of the mouse feet/mouse pad contact. Take the mouse, move it around quickly in circles, and see if it glides well. Then move it left and right on a line or as if to draw a star, and see if the mouse tends to “stick” at the ends of the line or at the points of the star. It may not be easy to feel a difference, but if you experiment enough you will find some surfaces (like if you are using sandpaper as mouse pad for whatever reasons) which are obviously not suited for gaming.

Point two: If you have dirt/sweat/sand/dust/whatever bizarre substance between your mouse feet and the mouse pad, then obviously your performance will be affected since it is not just the mouse feet and mouse pad in contact anymore.

Point three: The “feet” on optical mice are, as far as I am aware, made of polytetrafluoroethylene (PTFE), or as more common name goes, Teflon (yes the same stuff that keeps eggs from sticking to frying pans). Look out for the thickness. From the cheap mice I have used, it seems that the biggest problem is not that the mouse feet are not slick enough, but rather they are too thin, or the groove they sit in are too deep so after a while the portion protruding gets worn out and you have the plastic body of the mouse in contact with the mouse pad. When that happens if you want to play games you will need to replace either the feet, or the whole mouse. There are such products (mouse feet replacement Teflon stickers) around, but generally they are made for certain popular gaming mouse models. Of course, I (and anyone who plays Osu! with pure mouse) abuse my mice way more than average users, so it may also be the case that your buttons/optical sensors have died long before your mouse feet.

Point four: Another duh point, when you are gaming don’t press down your mouse with all your strength as if there is a cockroach between your mouse and the mouse pad trying to escape . If you can’t control the mouse while lifting it a little, then at least keep your hand relaxed. Also, this is one of the reasons why heavy mice suck, it increases the normal reaction force, and hence the friction. In fact a while back when I was using a very heavy Diamondback clone I had to consciously lift it a little, and even then the loss in performance was extremely obvious.
 

Optical Sensor

Once upon a time, I believed that all optical sensors on mice were equal. I was so very wrong, it turned out that some are more equal than others, and there are also those which are outright horrible (and not equal).

Optical Sensor: DPI

Years back when I first tried out optical mice, I thought that 800DPI was more than enough for anything, and those people who bought anything above 1000DPI were all idiots. It was an understandable mistake though, firstly because I was comparing with ball mice (which I can’t imagine using now, but am still forced to use), secondly because the standard screen size then was 1024x768, and lastly because I didn’t know that turning up the mouse sensitivity in Windows beyond the middle notch badly affects the mouse accuracy (further elaboration later in the software section).

For now, just believe me when I say that your Windows cursor speed settings should always be at the middle notch. With that setting constant, change your screen resolution down as low as you can, then move your mouse around to feel for the cursor speed, then change your screen resolution to as high as you can and do the same experiment. You should find that your cursor moves much more slowly when you are using a higher resolution. The reason is simple. Given any mouse on a fixed DPI setting, it will move the cursor a certain fixed number of pixels for each centimeter you physically move the mouse. If you use a bigger screen, this number remains the same; however it is now a smaller fraction of your total screen area. Noting that objects in games such as units in RTSes usually scale up in size together with the screen size, and that the number of pixels your cursor moves doesn’t, you may not be able to move your mouse across the screen in one clean swipe in a big screen. Lifting the mouse introduces inaccuracies, and should be something left to office workers using ball mice.

So as long as the DPI number is big and fat, the mouse would be awesome, right? Not exactly, it turns out that quite a few of the random mice I have tried (all with big fat DPIs) have problems with their optical sensors. For example, the cursor is simply not accurate, or it jumps around once in a while, or the mouse doesn’t work on a significant number of surfaces. It may not be obvious when you play RTSes (especially so when you don’t have another mouse for comparison), but it is a solid wall that stops your APM from improving any further (you lose the cursor too often or you have to do a second readjustment after trying to point to something, etc). You can open up Microsoft paint and try drawing circles. If you find that it is very difficult to draw smooth circles, or if the cursor jumps around periodically, chances are the optical sensor is not as good as the packaging boasts it to be.
 
Optical Sensor: Mouse Refresh Rate

Take your mouse for a spin (literally). Open up Microsoft paint; draw big circles as quickly as you can. Depending on the refresh rate of your mouse, the curve of your circle may either be smooth or have kinks (see below diagram). Standard mice have a refresh rate of 125 Hz, meaning that the mouse updates the computer on where it’s moving 125 times every second. If you can draw 10 circles a second, then your mouse only has time to update around 12 times while you draw 1 circle, resulting in around 12 kinks.

http://i286.photobucket.com/albums/ll109/p06513w/untitled.png

Now I can’t say for certain whether a high refresh rate makes a different in your Starcraft 2 or DotA games, but it does make a difference in Osu!, especially if the frame rate and the refresh rate of your mouse are both low.

Most gaming mice support higher refresh rates than 125 Hz (will usually be advertised in big words on the packaging), but even for mice that don’t there is a way of increasing the refresh rate through software methods. However, whether or not the mouse can deliver the higher number of updates is another issue. I tried it on one of my cheap mice before, and could only get it to around 250Hz.


Mouse Buttons

I’m not very particular about mouse buttons; though there are people who prefer their mice to have 10 buttons, all reprogrammable so that they can do things without touching the keyboard. I only use the right click button, left click button, and the scroll wheel. Great grand total of 3.

There are nevertheless a few things to take note, if you play Osu! like I do. Firstly, the mouse click shouldn’t be too “deep”, requiring you to take too much effort to click. This is especially important if your finger is short and your fingertips are not at the far side of the button. For certain mouse button designs, you need a lot more effort to click at the middle of the button where it is closer to the fulcrum of the button. The mouse click depth should also not be too shallow, and would preferably be deep enough that there is enough feedback for you to know that you have clicked it. (So that you know for sure you have clicked even though you can’t hear the click sound because you are wearing headphones)

I average up to a million (EDIT: actually should be closer to around 200k, a million is grossly over estimated for most typical weeks) clicks a week just from Osu!, so it is also important for the mouse buttons to be durable.

 
Ergonomics (weight and shape)

There are some huge, heavy and cool looking gaming mice out there which look absolutely bad ass. Guess what, chances are, a lot of these mice are designed for bigger and stronger Caucasian hands. A simple application of Newton’s law of motion F=ma would tell you that with a more massive mouse, you would need a greater force to achieve the same acceleration.

As for the size and shape, it depends not only on the size of your hand, but also the type of grip you use. I’ve seen quite lengthy descriptions of the different kinds of grips such as palm grip, fingertip grip and claw grip; you may want to Google them up. Feel for yourself the size and shape, but just keep in mind that lighter is usually better, as mentioned previously.