There's a scroll bar at the bottom of the sheet, being in 1280*960 res forces me to scroll right to change glyphs and spec.
If you don't pay attention after doing it you will think "all colours are gone, and all of my stuff is gone" while you are actually looking at the "Optimal" column and not the "Current" column. You only have to scroll back to the left :o
Happened to me a couple times before i figured it out.

I can confirm I had the same bug happen twice. Closing and reopening did fix it one time and the second time it was mistakenly saved after bug occurred before I was aware. I did not post yet as I am unaware of the specific sheet change I preformed that caused it each time, but I do believe it was in relation to a spec and/or glyph adjustment.

Yes, as having 735 ArP was soft capping me when using Runestone. I'll upload my exact file when I arrive back home, but I am back at work again unfortunately.

To clarify, since people seem to be having a problem here, this was the original comment to which I responded:

This is wrong as both myself and Aldriana have pointed out. It's really that simple and I'm going to stop any further comments about this topic.

This is a good point except it only applies to a spreadsheet based on average values. In the simsheet by contrast, variance plays a big role. To take your example with hit and expertise, low levels of hit increase variance in energy income and can lead to energy draught situations or on the other hand bursts of energy which can lead to rotation disruption or changes substantial enough to say they are not well captured with a spreadsheet based on average values. Similarly, unlucky dodges of envenom can lead to cycle disruptions for mutilate spec, just like in a real fight, and it's all captured in the simsheet.

Averaging in a simulation is done after observing all those trials, some of which may have suffered from cycle disruptions due to variance, and some of which may have had a lucky string of crits. Variance has a much more pronounced negative effect on DPS in a simsheet, just like in real combat.

To give you another example, when I was modeling HAT before, an obvious parameter was the average combo points per second you receive from group. However I also discovered that to a lesser extent, it mattered which class was in the same group with you. Some classes provided a more stable income of combo points (think low variance due to high frequency of attacks) compared to others (big delays between attacks, spikes in combo point income). Holding average combo point income constant, those classes that gave you a smoother income of combo points were more beneficial for your dps, and simsheet is the best way for capturing that second order effect.

To be certain, average-based spreadsheets are also capable of modeling these second-order effects, however they typically do not probably because these effects are deemed insignificant and are also harder to model. So basically the bang for the buck is not there. In the simsheet, I do not have to add any additional modeling to capture second or third order effects, it's in the nature of the method to capture all of them by simulating the RNG rolls the same way they are done by the game itself.

I think tetracycloide's critique runs quite a bit deeper than that Mavanas: you're still thinking about the effect of variance on average DPS rather than the effect of a variance in DPS on the raid outcome. As an extreme example, a setup that 90% of the time produces 9k dps and 10% of the time produces 1k dps (avg dps = 8.2k) in a burn phase just isn't as good as a setup that always produces 8k dps if you want to consistently 1shot bosses like twins or jaraxxus. Obviously none of the setups we're talking about here are nearly as extreme, and it probably matters more in 10s than 25s, but it's something we don't often seem to think about in our pursuit of average DPS.

After I banged my head repeadely against the table I can confirm this.

I opened one of my old "bugged" copies, scrolled all the way to the left and voila. Damn my laptops low resolution.

Thanks Hihu and I apologize for clogging up the forum with my stupid mistake.

He's talking about variance on a smaller granularity though, and admittedly this is pretty nitpicky, but the point was that hit and expertise reduce your infight variance, which could be significant - if a Mutilate Rogue got a dodged Envenom and dropped SND in the middle of a Jaraxxus Portal, it could (theoretically) be the difference between 1 spawn and 2. This gives them some non-measurable amount of "non-dps" utility.

We already know though if you are Mutilate and getting dodged Envenoms that cause SnD to drop, that you need to get more Expertise. You aren't in the mindset "I need more Expertise because I want to lower my variance" it's "I need more Expertise because I don't have a stable rotation."

I think the variance in DPS for Combat rogues is so little that not until you start reaching BiS gear for T9 that Expertise is a good alternative to more Agility, as Aldriana pointed out earlier in the thread.

Weren't there also discussions about Double WP not actually maintaining the Savage Combat Debuff?
It was posted just before the most recent Mut beta Spreadsheet and I think it got drowned out (understandably) by questions regarding that.

If duel WP doesn't maintain savage combat without another class in the raid able to do so, I don't see how it can be viable. Any other ideas on this? I haven't gotten around to testing it myself

I have a question regarding trinkets in the combat 1.3 (beta4) spreadsheet. Are the effects (click or proc) averaged over time for all of them? From what I can see they are, but i haven't dug deep into the calculations. I'm thinking about the additional value of a click-effect used with cds like BF, AR and KS. Right now the spreadsheet is telling me DMC-greatness is 14 dps better than Mark of supremacy, but I'm thinking Mark of supremacy might come out ahead in-game if used with those cds.