My Nvidia Story

[Page 2] Part 2




The Guide Delays Begin



Back to my tale. At Nvidia's suggestion, given games were again starting to be a bit more tweakable, my game guides for GeForce.com started becoming more detailed towards the end of 2011. Developer support also improved, such that for some games if I really got stuck working something out, I could ask the developers questions. Ominously, developer interest in the guides also began to increase, as the guides gained more prominence on GeForce.com.


The first of the more detailed guides I wrote for Nvidia was the Battlefield 3 Tweak Guide, and that was fortunately a relatively quick affair. I started the guide on 27 October 2011, and had it finished by 4 November, and released by 9 November. This only served to make me optimistic that we were truly in the groove now in terms of getting out decent guides on time.


But the next detailed guide I wrote, the Skyrim Tweak Guide, is when I started experiencing major delays.


I usually start writing a guide on the day of the game's retail release. On a couple of occasions we tried using pre-release versions of a game, but the problem is that most games change, sometimes quite a bit, in the weeks leading up to their release. Feature alterations, additions or removals, and substantial performance optimization can occur right up until the day of game release. Some games even receive patches on release day, incorporating even more changes at the last minute. The point is that it is too risky to do a game tweak guide based on a pre-release version of a game, on the rare occasions I had access to them, as the performance and tweakability of the game is still in a state of flux, and everything needs to be retested after launch anyway.


So I began writing the Skyrim Tweak Guide on 11 November 2011, the day of its public release. The process involved playing the game for at least a day or so to get a feel for the engine and its features, and to rack up a range of different save points encompassing various environments in the game. Then came the testing, researching, testing and more testing, and documentation process, ending on 21 November when I sent the final draft to Nvidia, complete with all of the guide images, for approval. The guide weighed in at over 15,000 words, so it was no small exercise for me.


Unfortunately, at this point, it got bogged down for two weeks before it was formally released on 6 December 2011. The bulk of the delay came from an incredibly slow turnaround time in getting formal developer approval. I kept in constant communication with Nvidia, explaining to them the importance of getting the guide out ASAP. When the guide came back from the developer, only a few sentences had changed, mainly in the introduction text, and it was slapped with a "Bethesda Recommended!" logo. Two weeks' delay for a time-critical Internet article, just to keep the developer happy?



Interlude 3: Developer Relations


As a GPU manufacturer, Nvidia's fortunes are linked closely to gaming. Therefore, Nvidia quite wisely maintains very close relationships with the major games developers and publishers. This partnership is actually one of the reasons why I've recommended Nvidia graphics cards for many years. When a developer has the chance to create and optimize their game on Nvidia GPUs, with close Nvidia tech support, and in turn, Nvidia can optimize their drivers with developer input, the end result is that when a game comes out, it usually plays quite well on Nvidia hardware from day one. This is particularly true for titles that Nvidia has sponsored with its The Way It's Meant To Be Played program.


Historically, Nvidia's main competitor AMD (formerly ATI), has had major issues in this department, with several cases of problems with games on AMD hardware shortly after release. However, in a surprising turn of events, AMD's Gaming Evolved program has suddenly gained traction in the past few months. AMD has managed to snag some major titles under its sponsorship banner, including BioShock Infinite, Tomb Raider, Far Cry 3 and Crysis 3. I'm happy about this as a gamer, but as we'll see shortly, it caused major problems in my work for Nvidia.


In any case, you can see that there is a symbiotic relationship between games developers and graphics card manufacturers. This relationship is very important to them, and frankly, it should be important to us as gamers as well. The closer these guys work together, the better the gaming experience for all of us. Unfortunately, it typically involves exclusive marketing rights and closer relationships only for the highest bidder, so either Nvidia or AMD users, not both, usually reap the greatest benefits in each case.



You Suck, Sellout!


While at first a few weeks of delay in releasing a guide may not seem like a big deal, any gamer can tell you that interest in a game peaks during the first week or two after game release. That's when most people who want to play the game actually need the tweaking information. A few weeks delay can mean the difference between someone productively using the information in a guide, and not needing it anymore because they already finished the game. People don't care why something is late, all they know is that it's not there when they need it.


To make matters worse, during the delay for the Skyrim Tweak Guide, people were steadily building up the guide to be much more than I could possibly deliver. It was, apparently, going to resolve every problem in Skyrim and boost performance by 300%. None of this hype came from me or Nvidia, in fact I tried to hose it down as much as I could, explaining that there were no magic tweaks in the guide. There is, after all, only so much I can do using the variables and settings available in the game within a week or so after its release. But because of the hype and the delays, I started getting wonderful emails telling me I sucked, and amazingly supportive posts on forums such as this one, calling me a sellout. When the guide finally did arrive, it was generally well received, but it had no hope of living up to the hyped-up expectations.


At this point, I had a long phone conversation with my Nvidia contact, and we agreed that the priority was to get the guides out earlier. I explained that at best I could only shave a day or two off my process, meaning I could get them a fairly polished draft of a guide within 5-7 days after game release at best. That's not too shabby, considering each guide can be 10,000-15,000 words or more, and requires a lot of research, testing and appropriate screenshots. The screenshots in particular are often incredibly time-consuming, because aside from having to find the right area to clearly demonstrate what are sometimes very subtle changes in visuals, with checkpoint saves the problem becomes much worse. If you have to restart the game engine to implement a tweak, loading up a saved checkpoint means having to get to the exact same location as the previous screenshot, and then lining up the scene for an identical comparative screenshot. This can be quite tricky.



From Bad to Worse


I did several more guides over the course of 2012, and the delays started getting worse, instead of improving as I'd been assured they would.


The Max Payne 3 Tweak Guide for example, which I started on 1 June 2012, actually wasn't completed until 15 June, because close developer involvement meant several changes to the game based on our feedback, which in turn meant waiting for them to implement the changes in a patch, then retesting and rewriting the guide to reflect the changes. This in itself isn't too bad, since we were helping to improve the game, so fair enough.


But once the guide was complete and sent in for developer sign-off, we were then forced to sit and wait as the developer delayed approval of the guide until it could get the next game patch out. The guide went from being ready for intended release on 21 June, to finally being officially released on 19 July - a full month of delay just for approval.


I'm all for close developer relationships, and I completely understand both Nvidia's position, and those of the game developer. But what's the point of a game guide that comes out over 6 weeks after a game? We could just as easily have released it within the first 10 days, then updated it twice with the relevant changes - which were not that major by the way.


While Nvidia accepted the validity of my complaints, in the name of maintaining developer relationships they said they couldn't do away with the developer approval process. They promised to speed things up though, and with some wishful thinking on my part, I accepted this.



Far Cry 3 and Crysis 3


Finally, we reach the Far Cry 3 Tweak Guide debacle. I was asked to do the guide by Nvidia, and assured that since FC3 was an AMD-sponsored game, Nvidia wouldn't need to get involved in a lengthy developer approval process. I began the guide on its release day in Australia (29 November 2012), and worked non-stop to get it done. The final draft was ready on 6 December and submitted to Nvidia. The guide weighed in at 13,000 words and had 106 individual images. I was very happy with the way it had turned out, and Nvidia also appeared to be as well. It was formatted for publication on GeForce.com the next day.


Then that old demon "developer relationships" sprang up to bite me on the ass again. It turns out that Nvidia did need developer approval for the guide, and thus now had to submit the guide for approval. Things become vague soon after this point, as my main contact person at Nvidia had changed by now, and the new guy was less than clear in his communication with me. Ultimately, it appears as though the developer simply withheld giving approval. The guide was dead in the water, never to be released it seems. I did get paid for the guide, which helped soften the blow, but also prevented me from publishing the guide myself, as Nvidia now owned the content. I was incredibly annoyed at having put so much effort into writing something that may have been of some use to readers, only to have it disappear. If I'd wanted to do work that virtually no-one ever read, just to get paid, I would have stayed in my old job as a government Economist and made considerably more money with much less effort and frustration.


Now here's where the expression "fool me once shame on you, fool me twice, shame on me" comes into play. Despite the growing delays and lack of communication from Nvidia, and one of my guides going MIA, I stupidly accepted the commission to do a Crysis 3 Tweak Guide for them. At this point, I'd been repeatedly assured that the Far Cry 3 guide fiasco was a one-off, and that we could work out some sort of alternate arrangement to publish the guides in a timely manner. Just prior to starting the Crysis 3 tweak guide, I was told that Nvidia could arrange alternate sponsorship via one of their board partners, like ASUS or EVGA, and thus get around the developer delay dilemma. It wouldn't be published on GeForce.com, but it would be published in a similar format for free somewhere else.


I started on the Crysis 3 Tweak Guide on release day in Australia (21 February 2013), and wrapped up the 15,500 word guide, along with its 130 or so images on 2 March. During this time, I was told the arrangement had changed, and the guide would be published on a major tech website instead. I accepted this as it was too late to do anything about it, and to sweeten the deal, Nvidia said I could also publish the guide on my own site at the same time. By 4 March 2013 I had formatted the guide and placed it on my own site, ready for public release as soon as it was published on the other site.


Then I started to see that things had gone very wrong. I was told on several different occasions between 5 March and 20 March that the guide was to be published "soon" on a particular site I won't name. It's one of the biggest tech sites in the UK, but one I don't have much respect for, as it's more of a tech tabloid focusing on smartphones, tablets, consoles and general consumer tech - hardly the PC gaming audience I wanted for the guide. But at this point, I just wanted it published.


It became apparent that whatever arrangement that Nvidia had supposedly had in place prior to completion of the guide actually hadn't been formalized, and they were making it up as they went along. I won't necessarily say that I had been lied to, but I had certainly not been told the whole truth. I kept asking for answers, but got the same message about it being published "soon", "later today" or "tomorrow" - all of which never eventuated.


On 16 March 2013 I emailed Nvidia to politely tell them that I couldn't keep writing guides for them anymore. I also publicly announced on my own site that the relationship had ended with Nvidia. The repeated delays in releasing the guides, which dated back to late 2011, were only getting worse. It was making the guides almost totally redundant given how late they were being released. It was negatively affecting my reputation, as most people could only assume the delays were due to me being slow with writing the guides, or that I was h