Blog: Volcano & sapphiRe in Seattle for CS:GO
Heather "sapphiRe" Mumm on Fri, 01/27/2012 10:54PM
Several months ago, Counter-Strike legend Sal "Volcano" Garozzo was invited to the Intel Extreme Masters event in New York specifically to meet with with developers involved in the Counter-Strike: Global Offensive project. Given Sal's unique experience as a player that competed at the top level of Counter-Strike 1.6, Counter-Strike: Source and even a brief stint with Counter-Strike: Condition Zero, in addition to his experience in level design and programming, Sal hoped to provide valuable feedback as CS:GO evolved. Consequently, Sal was eager to get in contact with Valve to initiative the discussions.
Since his initial meeting with Valve in New York, Sal has remained actively engaged with the game's developers. Since CS:GO was announced, Valve made it publicly aware that competitive player feedback is encouraged and welcomed. Consequently, Sal saw it as an opportunity to provide weekly feedback throughout the closed beta phase. Sal made sure to get contact emails and continued to send his thoughts over to Valve after playing the beta every day for a month.
As a result of his efforts to relay feedback to the developer team, Sal was invited to Seattle to demonstrate some of his suggestions on-site and to provide additional recommendations for CS:GO.
I had also tagged along to the Intel Extreme Masters event and was able to get a first hand look at the game. Valve was kind enough to invite me along on the Seattle trip. Yesterday, Sal and I began our journey to the West Coast.
Found in the Valve lobby
Upon arriving at the Valve Headquarters, located east of Seattle in Bellevue, Sal and I felt we walked into nerd heaven. The Valve lobby was like a gaming museum, with figurines from Portal on display, custom Left 4 Dead and Half Life computer cases lining the room, and a montage to Aperture Laboratories and Gordon Freeman shown on the lobby's TV. Located in the center of the room was a giant Valve.
We met up with Chet Faliszek who escorted us up to the CS:GO floor of the building. For those unfamiliar, Chet is a writer for Valve. He has been responsible for creating the story for Half Life 2: Episode One & Episode Two, Portal 1 & 2, and Left 4 Dead. His resume speaks volumes.
Chet has been closely involved with the CS:GO project since its inception.
As two people that work in the cube-style, corporate world, Sal and I were amused by the even the most minor of elements within the Valve office. It is a very collaborative style layout, with very modern furnishings. The desks are all on wheels and the heights are adjustable. This allows employees to move their desks to other areas of the office with ease. Everyone has a minimum of two monitors, but most have three or four. A number of employees work standing up for part of the day, then lower the desk to sit for the remaining hours. Many employees use medicine balls for chairs, others use $1,000 Herman Miller Aeron chairs. Every employee's desk has some sort of Valve memorabilia on it, from a demoman figurine to a companion cube and even a blow-up P-body.
We were given a quick tour of the floor. Custom game character paintings lined the walls. A classroom was located just down the hall where 20 computers were set up for students that came into the office to learn various game-related topics. A private bathroom with a shower was nearby for those long nights or after a trip to the fitness center, located on another floor. The lunch room was equipped with a plentiful supply of snacks.
After our tour, we headed out to lunch. We were joined by a number of Chet's coworkers. One was Mike Belzer who we had met in New York as well. Mike also has a rich history as someone that worked at the Disney Animation Studios and contributed to a number of Pixar films. Previously, he was involved in animating Gumby and The Nightmare Before Christmas, among with other notable films. Now he's working on animation at Valve. The lunch trip also included several level designers, animator, architects, statisticians, and artists, among others. It felt mesmerizing to be surrounded by such talented people that are involved in the game that I've loved for so many years.
Following a fancy lunch, we headed back to Valve where Sal and I were given two computers to set-up on. We were eager to get our hands on an updated version of CS:GO, different from the one we've been playing for the past two months.
We were set-up with our temporary custom steam accounts (with thousands upon thousands of games!) thanks to Ido Magal. Ido is a lead on the CS:GO project. He has worked closely with a number of professional players, including Sal, during their time in New York.
Sal and I had come with a lengthy
list of changes, suggestions and modifications we'd like to see based on our time in the closed beta. Sal also brought with him a list of the top three ideas from ten CS 1.6 & CS:S professionals. Some were absolute must haves, others were merely minor suggestions, and a few were some radical ideas (You can blame/thank da_bears if a "drugbust" mode gets implemented into CS:GO but I wouldn't count on that happening!). When designing a game for both competitive players and the masses, there is always going to be some conflict of interest in various decisions so while we didn't expect every suggestion to be implemented, we were very hopeful about a handful of ideas given the fact that Valve has been quite receptive to feedback thus far. More on that to come but just wanted to quickly note how impressed I was with how much data, both quantitative and qualitative, Valve has been collecting to analyze and utilize as the game evolves.
Immediately, we were able to at least cross off a few items from our list once we started, such as stamina related to boosting and an improvement to the scoreboard interface to allow for easier reading. We only had about 30 minutes to play since we were scheduled to head over to Hidden Path Entertainment for the remainder of the afternoon. Instead of going through our entire list of ideas, we were distracted by the desire to check out the CS:GO version of several classic maps.
We went into my personal favorite, de_inferno. There are two changes on the map that cut off angles and peeks that previously existed in CS:Source but didn't necessarily harm strategy. One change makes the map more like the CS 1.6 version. The gameplay should be nearly identical for CS:S players. For CS 1.6 players that are concerned about "clutter" in the CS:S version, de_inferno in CS:GO actually feels a bit more simplified. I couldn't find any dynamic props such as the barrel top in CS:S that gets kicked around in banana. The props that did exist were static. Overall, I would say the new version is a happy medium for both types of CS players.
Hidden Path Entertainment
At 3:00pm, our driver picked us up and took us a few miles away to Hidden Path Entertainment. HPE was contracted by Valve to work on CS:Source and now they've been kept on-board for CS:GO. We met up with Steve Kramer and Peter Freese. Steve is working on CS:GO gameplay and Peter is the programmer.
We brought along our checklist of ideas from both ourselves and other CS 1.6 & CS:S professionals. We entered into a nearly four hour long conversation about CS:GO, CS:S and even CS 1.6. Sal did most of the talking since he's the professional, but I chimed in every now and then.
It was an absolutely eye-opening experience. As someone whose programming knowledge is very basic, I was amazed to learn about all the technicalities of game design. Steve said one thing that really resonated with me.... His most powerful tool as a game designer is Microsoft Excel. Every part of the game really comes down to numbers, graphs and statistical models. Though Steve said one other thing that I'm sure every competitive player can appreciate. Even though every change is driven by numbers, "feel" is just as important. You can change a value that would numerically make sense, but does that feel
right? That's where we come in. That's where you come in.
And that brought us into a discussion about player feedback. Everything comes down to a number and then testing that number to see how it feels.
One of my biggest pet peeves throughout the closed beta phase has been reading player feedback that sounds like this.
"I don't like it. It needs to be more like CS 1.6."
"I don't like it. It needs to be more like CS:S."
Well, what needs to be more like CS 1.6? The recoil? Ok. What about the recoil? Is it really
the recoil that is "wrong"? Or is it because you're having trouble seeing bullet impacts so you're not able to control your recoil? Or is there a different issue at play?
I was able to dig into the numbers behind recoil. It is simply just a pattern in number format. Even in CS:GO. No different than CS 1.6 or CS:S. Maybe the pattern has changed, but it's still just a repeating pattern that needs to be learned. One recoil pattern was learned by CS players in 2000. Another pattern was learned by CS:S players in 2005. Is it the right pattern? That I'm not so sure about yet and I'm sure it will be debated by many. Does it feel right? That is another question to ask. My personal feeling is that it could be tightened up a bit but I'm sure everyone has their own wishes. What I am sure of is that feedback should be provided throughout the beta phase, both good and bad, from all participants regarding all elements of the game.
After discussing feedback with both Valve and HPE, they reinforced the importance of eloquently sharing feedback and I'd like to extend the same thoughts to you. Constructive
feedback that is actionable
is far superior to a list of ranting.
Valve and HPE were both thrilled to welcome feedback that is informed, realistic, constructive and actionable.
Hidden Path Entertainment's lobby
Anyways... back in the HPE office, we were shown several graphs on a markerboard to explain grenade damage, base inaccuracy, hitboxes and weapon animation not only in CS:GO but in other Counter-Strike games. Despite playing this game for 10+ years, I never put a number behind the strategies I use. But numbers exist. In a way, I had hoped everyone had this opportunity to dig into the details behind the game. It can only improve ones gameplay style to understand exactly how, for example, grenade damage is distributed. I was never more excited to get back to a computer and play CS than I was that day.
After a lengthy, but informative discussion of Counter-Strike, we took a quick tour around the HPE office. Once again, we were greeted by a very collaborative, relaxed atmosphere. One room had an entire wall full of professional artwork. The paintings were of Counter-Strike player models that didn't make the cut into the game. The refrigerator was custom made. It was named "The Defense Fridge", a play on "Defense Grid", one of HPE's most famous titles.
The lobby was lined with game boxes from all the shipped titles that involved HPE. We spent another thirty minutes talking to Steve. It was so refreshing to see how passionate he was about game design.
When I'm sitting at home behind my computer, it's hard to realize what truly goes into making this game I've spent thousands upon thousands of hours playing. I forget that there are passionate people over at Valve and HPE that practically pour their soul into it. It was around 7:30pm that evening. I, along with many, would have headed home by now after the work day. But the employees at Valve and HPE were still hard at work and it seemed like they were there by choice. I would imagine their work hours ended at least an hour ago, but none showed any signs of complaining. The usual office would have someone saying "Is it 5:00pm yet?" or "This has been a looooooong day." I never once witnessed that at Valve & HPE.
Blog: Day 2 at Valve
Heather "sapphiRe" Mumm on Sat, 01/28/2012 8:43PM
Our first day at the Valve headquarters was spent mostly being introduced to various people involved in the CS:GO project. The plan for Day 2 was to go through the list of painpoints and suggestions Sal had prepared, along with the list from other CS 1.6 & CS:S professionals.
We arrived at Valve around 11:00 AM. We weren't able to get into the beta right away because it was in the process of being updated. So instead, we headed into a conference room with several Valve employees to chat.
Sal had prepared eight pages of bulleted thoughts that we discussed over several hours. Those fearing Sal is "too nice" to share his discrepancies would surely be proven incorrect here if they were a fly on the wall. Valve had asked him outfront not to hold back any ideas or concerns which Sal had planned to do from the get-go regardless.
A condensed list of the items up for discussion included controlling recoil, footstep radius and volume, ambient noise, dust sprites and ambiance, HE grenade damage and armor penetration, AWP delay, grenade cooldown, HUD text & UI, competitive vs. casual mode differentiation, friendly fire damage, among other items. (Sal's entire list of suggestions will be withheld for a future blog that he will author.)
One note to point out is that this list was created prior to us playing the updates Valve had been applying to the development version of the beta so some didn't necessarily apply anymore as they had already been addressed.
Every idea we brought to the table was written down on a notepad by the developers. Some suggestions were met with immediate agreement and we were told they they were already on or will be added the "to do" list. Other ideas required some debate and compromise. Then there were more ideas that were put on the backburner as they required more data. As I had mentioned in my Day One blog, Valve collects massive amount of data, both quantitative (numbers) and qualitative (player feedback).
Anyone familiar with statistics will know that a larger sample size is generally optimal to improve your confidence index, hence assuring the decisions you are making are shared with the sample population. This is where we told the broader beta phase will come into play. The first beta allowed access to a very small population and was primarily focused on testing game stability. When the beta opens to a broader audience, that is where Valve will watch the numbers and watch the feedback to determine next steps. It was enormously interesting to learn about the entire process and dig through some of the numbers.
Valve really wasn't kidding when they made that tweet. They literally analyze every bullet fired and any possible metric you could gather from that data. At least for me, since I'm a qualitative and quantitative data analyst by trade, I was blown away by the entire process.
Since the closed beta was for monitoring stability rather than responding to feedback on a whim, it will be interesting to watch how both the quantitative and qualitative feedback collected through the broader beta phase will drive decision making for future updates. Learning Valve's entire process for collecting data and feedback, analyzing the results and creating a plan was incredibly interesting. It gave me a sense of hope to know so much thought goes into even the most minor of decisions and moreover, that the competitive communities feedback is heard and welcomed.
One of the many Microsoft buildings
Following our meeting in the conference room, we went back to our computers to check if the update was complete. Since we were still waiting on a few additional changes, we had decided it was a good time for lunch. Yet again, we were treated to a five-star steakhouse, located within one of the many Microsoft buildings. We were joined by Brian, one of Valve's many psychologists on staff
. Yeah, that's right. A Psychologist. I was just as surprised. Brian's profile is not listed on the Valve website but to reference another employees bio, here is what a psychologist does at a gaming studio. "Applies both psychological knowledge and methodologies to game design. Essentially this means he gets to play with data, perform research, and act as an in-house consultant of sorts." Here is another quote from Valve's career page regarding their psychologist position. A psychologist "design(s) experiments to evaluate various gameplay hypotheses and design choices."
Even if Math or Psychology were never your favorite subjects, a Counter-Strike fan can still appreciate a discussion with Brian. This was one of our favorite parts of the trip. When Brian demonstrated how data can explain and predict human behavior in a video game, we were blown away. He showed us data and a hypothesis about strategy on a particular map. We explained how game play actually works in competitive play. The two were nearly perfectly aligned. Unfortunately, it isn't my place to go into the fine details but it was just another realization of how truly intelligent these people are.
After geeking out with numbers, Sal and I headed back to our desks, plugged in our gaming gear and got to work.
One of the things that drove us crazy in the beta that is currently being played is the loss of stamina when boosting. For example, try boosting players onto catwalk from CT spawn. It is enormously frustrating. This is because a player was losing stamina for jumping onto objects. That issue was thankfully resolved based on community feedback so boosting felt no different than in CS 1.6 and CS:S.
Another beta element that bothered me was the scoreboard UI. It was difficult to hit tab and quickly absorb information. I died one too many times with my scoreboard left up as I was trying to understand who had the bomb, who was alive and who was dead. That should be less of a problem with the new version. It still has a "money" column, which I hope remains a permanent change. A player's "score" is still there, which I think for competitive player isn't too useful, but at least in this new iteration, it's far enough away from a player's F
ratio and money that it doesn't become overwhelming. The changes were a direct result of Volcano's feedback that he had emailed in prior to our meeting.
Volcano in the Valve lobby
Also, I started using the new crosshair rather than the legacy crosshair thanks to our discussion with HPE. Sal and I both immediately switched to the legacy crosshair in the beta because the extra bars felt distracting. After a lengthy discussion about how the additional bars show recoil based on weapon and based on player position (standing vs. crouching vs. running vs. walking), I promised to at least give it a try. After using the same crosshair for more than ten years, it took some getting used to but has allowed me better control in the game's current state. I'm glad to see that players have the choice of using either crosshair style based on preference. Allowing for the customization of various options is what I'd consider to be ideal to satisfy both the competitive and casual communities.
On our list of changes was a reworking of dust sprites, specifically on de_dust2. We didn't necessarily plan to insist they had to be removed, since we understand Valve wants to make a visual appealing game as well, but we had hoped dust sprites could be moved to areas where gun battles don't often happen. That was another change we could check off our list right away. This had already been changed based on feedback during the closed beta. For example, the middle area of de_dust2 was cleaned up. It was easier to see down middle from T-spawn, without dust sprites blurring vision. Some were removed entirely from the B bombsite as well and placed off to the side in the B halls. Overall, de_dust2 felt less "foggy" compared to the current version in CS:GO.
While running around on de_dust2, I had remembered an exploit
that Westerman had found. I showed this to the developers. They demonstrated how to submit a bug report. The bug was then flagged to someone involved in the map team and a resolution was already in the works.
Heading into de_inferno, I found another area where a player could boost to see through a wall in banana. Bug report submitted. That took Sal and I on a bug-submitting journey.
CS:S players will be happy to know I submitted several bug reports for exploits that have been the bane of my existence as an ESEA-League administrator. Examples include throwing flashes through the cement encasing on the roof of de_nuke, seeing shadows cast on ceilings, through walls and through doors, among others.
TF2 posters lining the walls
One issue in CS:S (and on de_dust2 in CS:GO) that can be very frustrating is when players can peek through tiny cracks between stacks of boxes and barrels within the actual model. It's not necessarily the act of seeing through that is the problem because it isn't necessarily an exploit that a crack exists. However, the knowledge you gain as the peeker compared to the limited perspective the opponent has can cause some imbalance.
For example, in CS:GO, a CT-player in the A bombsite can peek down long A through a crack within a stack of boxes to see with near perfection. The terrorist would be able to see the light contrast if a player was strafing in and out of the crack, but a CT model staying perfectly still would be difficult to identify unless the terrorist player had an AWP.
As a result, Volcano fought vigorously for a solution to this problem. He proposed two solutions that could work. Either boxes and barrels be pushed closer together to disallow gaps or the gaps should be slightly increased so players on both sides can gain similar knowledge. Valve seemed very welcoming to this idea following Volcano's persistence.
In CS 1.6, only a handful of boxes have tiny gaps in them compared to CS:S, but those boxes can be spammed so the trade-off is more balanced. It has always been a painpoint in CS:S so we hoped issue would head in a positive direction as CS:GO evolves.
For the remainder of the day, we went through an extensive review of de_dust2, de_dust, de_inferno, de_nuke, de_train and de_aztec to find any exploits that need to be resolved.
Once again, the Valve employees were all working diligently (and quietly) at their desks until 7:30pm when we headed out the door.
I was hoping to play my ESEA playoff match from the Valve offices, but decided it was best to head back to my hotel so I don't distract those working around me with my mouse slamming and play calling throughout the match.
Instead, I hustled back to the hotel, fired up my laptop and phone tether and played my match with a glorious 150-300 ping and 30 FPS. It was a disappointing end to another insightful and enjoyable day. But at least I was able to outfrag Wizard ( not like that isn't a standard already ;-] )