Have you ever been asked by a friend to help them move? If so, have you ever had the experience where, once you showed up to help, you realized that that person hadn't put in the time and effort beforehand to properly plan out how to utilize the time you had personally given up to be there? If so, then you might have experienced the frustration of finding yourself mostly standing around waiting on the person who was coordinating the move. If this is something you have experienced, then you are probably aware of the natural tendency to dwell upon the other things you could have been doing had you not committed to help your friend out. If this goes on long enough, it can even create feelings of resentment depending on what sacrifices you made in order to be there and help your friend move.
What does this have to do with game design? Well, designing a game is a lot like moving. It involves task management, time management, and coordiation - all of which are organizational skills. Even the most creative, artistic person will never get their game design off the ground without some kind of organization to the process - especially when the process involves utilizing the willing time commitments of other people.
The long, tedious process of trial-and-error that is playtesting can try the patience of even the most committed designer - sometimes resulting in a designer having to leave a project for a while so as to reapproach it with a fresh perspective later on. However, the gracious people in our lives who are willing to help us play-test our games are, in most cases, nowhere near as committed as we are to the process. How can they be? It's not their game design. They are simply helping us with our project. If one sits down with playtesters to playtest a game, it's usually with the understanding on everybody's part that the game could go on for a specific amount of time (perhaps a half hour to an hour for example). Thus, some time will have necessarily been set aside by all involved in order for a playtesting session to occur. If a designer has not done any solo playtesting of the design before asking for the help of others, then I assert that he is being disrespectful of the other playtesters' time as basic problems (such as turn-order sequence problems, basic ambiguities in rules, etc.) are problems that can often be discovered in simple solo playtesting sessions first before ever needing to bring other people into the process.
People will often feel more committed to a project if they have confidence in how the project is being managed and will often feel a greater sense of satisfaction with respect to their contributions if they see that their efforts are being utilized effectively. If your playtesters feel like their time and contributions were meaningful and well utilized, it's more likely that they will sit down to a future playtesting session. Solo playtesting in advance makes that more likely as one is more likely to "catch" many of the basic problems that would have come up in the playtesting session and, thus, can allow for playtesting sessions to become more effective and informative. Thus, solo playtesting is important and (I would argue) necessary in the process of designing a game.
However, there are also problems that solo playtesting can't solve. In fact, if one is doing too much solo playtesting without bringing other people into the process, a form of "myopia" can develope in the design process which, if allowed to persist for too long, can result in a lot of wasted time on the part of the designer and a lot of problems that group playtesting would have corrected. Let's look at three of these problems.
Often, a designer will spend so much time thinking about his game design that certain mechanics or rules seem more simple to him than they would to other people who haven't been spending hours and hours deliberating on how the game should be structured. I've experienced this before where a rule I felt was fairly simple was actually very counterintuitive or was simply more fiddly than I was giving it credit. Having the perspective of others to alert one to these aspects of a game's design is important.
Because solo playtesting requires one to think through and keep in mind what all of the players involved would have to consider with respect to their decisions, it can create the illusion that the choices facing the players in the game are actually more involved and deep than they really are. If a person settles down into the role of one of the players instead of trying to play as all of the players, one might find out very quickly that the choices facing any one player are actually not that interesting or, even worse, are overly simplistic.
When solo playtesting, I've found that it's pretty easy to make decisions within the game based on assumptions of value in certain paths, items, rewards, etc. within the game's design. Thus, solo playtesting sessions can create a self-fullfilling prophecy where the designer's assumptions about the values of different resources, objectives, or paths lead to decisions that reflect those assumptions and the game works in the solo playtests as a result. However, if other players sit down to play and don't have those same assumptions of value in mind, the game might actually not work at all. Having the perspective of others helps eliminate this tendency in the testing process.
So, solo playtesting is, from my perspective, both necessary and insufficient. It's necessary to engage in so that we don't waste other people's time or try their patience but it's also not sufficient as certain problems can only be solved via group playtesting sessions.