GUIDE - Deciding on a new mouse About a week ago, I started wondering if it would be a good idea to write this, so here goes! When buying a new mouse there are several things you have to keep in mind, such as mouse prediction, the speed until you no longer have perfect control, etc. I'll offer as much info as possible, but in the end a mouse is a personal thing, not everyone will like the same mouse and that's usually down to the shape. Mouse Prediction Mouse prediction is an algorithm that will try and predict the path you're trying to take. The problem with it, is that when making very small movements of just a few pixels, the mouse won't always make a perfect representation of your hand's movements. How can you test if your mouse has mouse prediction? Open up paint and just try and draw a straight line. Without prediction, the line will usually be imperfect, with prediction the mouse will try to correct the path and at a certain point the line will be perfectly straight. Another way to test it is by looking very close to your monitor until you start seeing the actual pixels (only works on LCD's, not on CRT's) and then try to move in a square of 4 pixels. Lowering your dpi helps for this test. Do I wan't mouse prediction? It really depends on the user. If you were using a mouse with mouse prediction up until this day and it never bothered you, don't worry to much about it. Another noticable thing is that mouse prediction can actually be a good thing on higher resolutions. There's more pixels for the algorithm to work with and it kind of helps to make your movements more stable. On lower resolutions however it's often better to have no prediction. List of mice without prediction: Razer Deathadder (Depending on firmware used) Logitech G500 (can be disabled/enabled through setpoint, logitech calls it "angle snapping") Logitech G9x (I'm assuming, since it's the same sensor as the G500/Xai) Steelseries Xai (Can be disabled/enabled through software or LCD) Pretty much all of the old MS Intelli mice Perfect Control, Meters/second, Malfunction speed Perfect control is a term that kind of came to life after the big ESReality Mousescore review done by Sujoy: http://www.esreality.com/?a=post&id=1265679 Instead of reviewing mice the old fashioned way, he would use a device to measure the maximum speed a mouse could handle before it quits on you and skips. Depending on how the mouse handles "skipping" it'll either just stop tracking or even worse, making you flip around and look at the sky or ground. This is probably THE MOST important thing about a mouse for a FPS gamer. It basically reflects how reliable a mouse is. How fast can you flick a mouse around until it quits on you? Laser mice, aside from the latest ones, usually have a very bad top speed, below 1.5 m/s. Opticals on the other hand can go above 2 m/s or even higher. This is where the MX518 and old Deathadder really start to shine. These are 2 mice you can throw around like a madman, they won't skip on you, not even when reaching 4 m/s. The newer laser mice such as the Xai and G500 have a much higher perfect control speed compared to the G5 and Ikari (and Copperhead). Though 1 detail should be mentioned. As you increase the DPI on these mice, the treshold for perfect control will start decreasing. But it's not really a big issue, unless you're one of those people using 5000 dpi and an ingame sens of 0.1. Sadly, mouse manufacturers still don't like putting this numbers in the specs, so in the end you're pretty much stuck to trying a mouse or reading comments on forums. Now on the graphs of Sujoy, aside from perfect control, there's still another "state" of tracking, usually where the graph stops increasing, but still keeps going. In this state, the mouse will have reached the maximum speed at which it can "move", but it is still capable of tracking. So moving your hand faster won't really translate into faster on screen movement. A nice example of this would be the MX518: http://www.esreality.com/?a=longpost&id=1265679&page=14 In the end, the most important thing to look at is the malfunction speed (when it skips), or better yet, if you get a chance to test a mouse, open up a game and swipe it across the mousepad as fast as you would possibly move it while gaming. Mice with a high malfunctioning speed (the higher the better): Steelseries Xai Zowie EC 1/2 Logitech G500 Logitech G9x (I haven't tested this one, but it has the same sensor as the G500/Xai) Basically all the Razer mice that were released since the Deathadder All of the Logitech MX series mice (518, 510, 300, ...) For other laser mice that aren't included in the list, if they use the AVAGO 9500 sensor (G9x, Xai, G500), their malfunctioning speed should be pretty high. Another list with more recent mice: http://wikis.jp/interfacedevice/index.php?MaximumSpeed_en How much DPI do I need & Sensitivity Your camera standpoint moves around in a circle of 360°. There's no real "best sensitivity" but there's 1 way to test if you could do with more dpi: To test, set your dpi quite low as it'll be easier Draw a line of 1 pixel thick in photoshop/paint Now get close to the monitor so you can actually SEE the pixels (only works on LCD's) Use the line as a guideline and try and move your mouse so that you can move in a perfect square of 4 pixels clock or counter clock wise Now you've got an idea of what the maximim precision possible is like on your monitor (obviously best done at native resolution). After all of this is done, go ingame (keep the low dpi) and try to make a very small circle, similar to this square. Your viewpoint moves in degrees, but we're looking for the maximum possible precision on your monitor. You can try a really high sensitivity like 10 ingame (in CSS, Quake) and you'll see that you're not capable of making very fine adjustments. After you understand this, start from your own sensitivity you had and see if you can make a very fine movement, similar in detail to the square in paint/photoshop. Lower your ingame sensitivity until you reach that point. This sensitivity is what you could consider the ideal one for your setup (monitor, etc), setting it lower then that will not give you a visible increase in precision (it's still there but you're limited by what your monitor can show you). Now all that's left to be done is comparing this sensitivity to your old one and increase the dpi you were using (not the low one for this test obviously, the one you used to play with). An example: going from sensitivity 2.7 to 0.9 with a dpi of 450: 2.7 / 0.9 = 3, so then you multiply your dpi with 3 450 * 3 = 1350 DPI This will give you the same sensitivity but with the ability to make more precise small movements. This also brings me to the point of the windows slider. Going above 6 out of 11 should be considered pure evil. 6/11 is ideal and going lower won't be a big problem, windows will simply use interpolation and discard inputs. Also the windows slider it's scaling is NOT what people think it is, 3/11 is NOT half of 6/11. The scaling also depends on wether you have "Enhance mouse pointer precision" ticked or not (turn that off btw, EPP = Mouseacceleration). With EPP ticked off: 1/11 = 0.03125 2/11 = 0.0625 3/11 = 0.25 4/11 = 0.5 5/11 = 0.75 6/11 = 1.0 For more see this thread from MarktheC: http://www.esreality.com/?a=post&id=1847344#pid1847344 When games use rawinput, which is better, windows input will be ignored and your sensitivity will be similar to what it is when the windows slider would be 6/11. If you're changing the windows sensitivity or going from 4/11 to raw input, you can simply calculate your new ingame sensitivity or new DPI like in this example: We have a mouse with 1800 dpi, sensitivity 1.0 ingame, 5/11 in windows and we would like to switch to rawinput or windows sensitivity 6/11: Simply multiply 1800 with the value for 5/11, 0.75 and this will give you your new DPI: 1800DPI x 0.75 = 1350DPI Same for the sensitivity ingame. We have sensitivity of 2.0 ingame and 4/11 in windows. We want to change the windows sensitivity to 6/11. Sens (2.0) x Windows 4/11 (0.5) = 1.0 ingame I also suggest you install a mousefix. Most games won't need it these days, but some older or badly coded games will force EPP on, giving you acceleration. A mousefix simply edits the coordinates of the acceleration so it reverts back to a straight lineair line. The best mousefix would be the WCAFix, which actual disables the code in windows for EPP: http://rapidshare.com/files/414877656/WCAFix_AllOS.rar On windows 7 and vista you will have to enable testsigning though. For registry mousefixes, 7 requires a different one then Vista/XP. Microsoft fixed a slight issue on 7 with how mousemovement was translated to on screen movement. For Windows 7 the best fix would be MarktheC's: http://www.esreality.com/?a=post&id=1846538 For XP/Vista the ideal fix is Cheese's, I still need to find a link though, the original one is down: http://www.esreality.com/?a=post&id=1548753 Shape The only tip I can give you after trying loads of mice: For god sakes, stop looking at the pretty lights, you're supposed to be able to hold that thing for several hours without getting cramps... Preffered shape is completely personal. Build Quality Right now we're not taking into account the few cases where people their mouse breaks quickly because of manufacturing flaws or w/e, we're looking at pure raw built quality. I've opened up and looked inside a ton of mice and if I where to pick a brand for the nr1 spot for build quality, it would be Logitech, especially the old MX series. All of the MX mice I ever had are still working, even right down to the 10 year old MX300... Razer seems to have improved, but their scroll wheels tend to break quite fast. Steelseries Ikari seemned to have lot's of flaws, I took a risk buying the Xai, it seems they didn't make the same mistake twice. I don't know about Roccat, if I had a bit more cash to spare, I would've bought the new Kone as well to see how it performed. A few more words of advice Until all reviewers start testing malfunctioning speeds or turn into fnatic f0rest, don't bother with mouse reviews. They're utterly POINTLESS. They'll keep sprouting the same crap over and over: "Feels nice and comfortable" "Looks nice" "Tracks very well" (with the reviewers mediocre gaming performance any mouse would tbh) And the list goes on. The best way is to do some research on forums like ESReality and Razerblueprints. Does a mouse have a problem that seems to be happening to more then half it's users (or in Razer's case, everyone...)? Then stay away from it. About Razer lately, I suggest staying away from mice from Razer with Z-Axis tracking at the moment, this is why: http://www.youtube.com/watch?v=wD0ThxpVHMs Extra Info A good, detailed list of mice and their used sensors and specs, made by skylit on the overclock.net forums: http://www.overclock.net/mice/854100-ocn-mouse-sensor-reference-performance-sheet.html EDIT: Windows 8.1 Fix This will fix the strange mouse behavior in some games (and the low polling rate reported in some mouse utilities). Link to Microsoft KB article: http://support.microsoft.com/kb/2908279 2 methods to get the fix: Through regular updates Install the hotfix (lost the url need to get this one back) How to apply the fix pretty much system wide: Open up regedit.exe Browse to "HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AppCompatFlags\" See if there's a ''Key" (like a folder if you don't know what a key is) named "Layers" If not present, create a new "Key" in "AppCompatFlags" Inside "Layers" create a new string value named C:\Windows\explorer.exe As value for this string use: NoDTToDITMouseBatch Once done, close regedit and open up a CMD window with admin rights execute the following command in the CMD window: Rundll32 apphelp.dll,ShimFlushCache Reboot just to make sure (registry adjustments were made in HKLM afterall) This will make the fix active whenever explorer.exe is active, so practically always. If you only wish to trigger the fix for certain applications, follow the same steps as above but replace the string value name with the path to your application (per example: X:\pathtoyourapp\csgo.exe). Repeat this for every needed application if you're not planning on doing this "system-wide" (using the explorer.exe trick). Example Once you've done this, your mouse works just as in Windows 8.0.